[fuoqi]

>you can specify the 2-vectors on the unit circle as "{ (x,y) : x,y in R | x^2 + y^2 = 1 }"

Which is not possible for any language targeting real machines since floats have fixed precision.

[icen]

You can specify the 2-vectors on the unit circle as "{ (x, y) : x, y in R | x^2 + y^2 - 1 < 0.000001 }" instead.