Re: Implement the "unum" representation in D ?

[bearophile]

Iain Buclaw:

Most encoding formats use a form of discrete cosine transform, which involves floating point maths.

I don't believe this much :-(

Bye,
bearophile