[MLton] Re: [MLton-commit] r5662

[Matthew William Cox]

On Thu, Jun 21, 2007 at 01:30:53PM -0400, Stephen Weeks wrote:
>  I wonder if it would be so bad to use bitwise equality for reals.

Speaking as someone who does some numerical analysis, I haven't used a
programming language's equality operator since I learned better. It's
somewhat of a surprise that the equality operator in this case is
actually somewhat intelligent.

I don't think that using bitwise equality will break any numerical code,
since in numerical code we're testing the absolute difference of the two
values against a tolerance anyways.

The people most likely to be affected are would-be numericists, such as
people experimenting with simple games or physics simulations. 

My $0.02
Matthew