Re: units library: A hash function approach type safe quantities

From: Eric Ford (eford_at_[hidden])
Date: 2001-10-02 11:15:24

> But what I want to avoid is assigning a variable X of type
> future_dollars/lbs_sugar variable to a variable Y of type
> current_dollars/lbs_sugar.  Both would be in units of "dollars/lbs"
> but are very different concepts.  BUT, I should be able to compute
X/Y
> and have it come out as future_dollars/current_dollars  (which would
> have units of pure).  Likewise, Y/X should be
> current_dollars/future_dollars.  X/Y should not interconvert with
Y/X
> inspite of both of them being of type pure. 

I beleive you are confusing units with dimensions.  Both X/Y and Y/X
would be dimensionless, not unitless.  There are plenty of
dimensionless things which need to be kept separate (e.g. radians and
number of apples).

E