big double?

[pakLebah]

Hi,

Any plan to support big double? Say, to count division up to n digit precision. You know, like digit of pi calculation but more natural.

Thank you.

-Bee

[lorentey]

Oh, definitely! With big integers already available, it seems it would not be too hard to build basic arithmetic operations for arbitrary precision fixed point and/or floating point numbers. Rational and interval arithmetic are also quite straightforward to implement, I think.

Implementing common transcendental functions for these types would be an interesting exercise. (I've never done that before. Then again, I haven't build a bigint lib before this one, either.)

I don't expect I'll have resources to work on these before the end of the year — it would be a nice weekend project, but I can't promise I'll do it soon. PRs are welcome. 😛