That's correct. My use of uint is not a perfect replacement for long and is based on assumptions made for a specific project. I should have been more clear about that.
ActionScript does not support long for portability purposes. --- In [email protected], Maciek Sakrejda <msakre...@...> wrote: > > >I prefer uint over Number for representing a long. > > To clarify, uint is just that: an unsigned 32-bit integer. A long is > (typically) a signed 64-bit integral number. A uint will allow you to > express numbers that are twice as large as the largest int, but that's > only a small fraction of the range of long (and it ignores the negative > range entirely). > > A Number, as mentioned before, is an IEEE-754 double (64-bit floating > point number). The problem with representing longs as double is that at > some point, you'll find a long x such that > > ((Number) x) == ((Number) x + 1) > > due to the properties of IEEE-754 floating point (i.e., there simply > aren't enough bits in the representation to distinguish those two). I'm > not sure what that long x is exactly. > > So basically, there is no long in ActionScript. You can approximate one > with uint, you can approximate one with Number, or you can write your > own BigInteger class and do the math internally. That's probably going > to be quite ugly and complicated, especially since you can't use > operator overloading for basic arithmetic. > -- > Maciek Sakrejda > Truviso, Inc. > http://www.truviso.com >
