C++ Big Math Data Types

[davef21370]

Say I have a value of several billion which I need to multiply by a tiny, tiny number then divide the result by another very large number (it's all to do with gravity) then do some adding up etc., what are the best data types to use ?
And is there a library for doing this kind of calculation without the loss in precision of floating point arithmetic ?

Any help appreciated.
Dave.

[-rst-]

Are you sure a 'long double' is not enough?

Maybe this http://gmplib.org/

[davef21370]

Oh yes indeedy -rst-, long doubles are plenty!!

If only I'd defined the gravitational constant as 6.67384e-11 instead of 6.67384e11, me stupid.

I do believe NASA crashed an otherwise perfectly good rocket by omitting a minus sign

Cheers.
Dave.