pow exponent type issue

[jmh530 via Digitalmars-d-learn]

I'm a little confused on why pow behaves so differently when 
switching from an int to a uint for the exponent.

import std.math : pow;
import std.stdio : writeln;
void main()
{

        float x = 2;
        int y1 = 1;
        uint y2 = 1;

        writeln(pow(x, -y1));  //prints 0.5
        writeln(pow(x, -y2));  //prints inf

}