Allow vectors that have norm(v) = 1

[MxmUrw]

Our current implementation only allows vectors where the [0,2^n)-space norm is encodable in [0,2^(2n-2)). This means if a norm is exactly 2^(2n-2) we cannot encode it, and thus do not allow it.