[VECTORS] Rotation without trigonometry

How is it done?
Normally, rotating a vector A(x1, y1) through t radians in order to get vector B(x2, y2) is done with
x2 = x1*cos(t) - y1*sin(t),
y2 = x1*sin(t) + y1*cos(t).

But trigonometry is expensive, especially without an FPU, so there must be another way.
I was thinking of something similar to Bresenham's circle algorithm, or perhaps reflecting the vector about the origin or something.

So, uh, how would you do it?

>>17
Off the top of my head I can't determine whether it is cheaper to flub FP math or do a bit of continued fraction arithmetic, and I'm not going to investigate. It is probably cheaper to flub FP, since CF arithmetic is basically matrix multiplication.