In article 
<Pine.LNX.4.21.0103181711090.22091-100000 / crimson.propagation.net>,
 ale / crimson.propagation.net wrote:

> First your code created a bunch of objects. Then it looped
> and measured the time to sum linearly one million integers. It was
> fast. It was fast because integers are kept in the object reference
> itself. (And Ruby just pretends to have them in a real objects.)
> 
> Calculating with floats was slower for couple of reasons. First each
> summation generated a new float object (new object instead of no object
> for integer).

Thanks for the explanation.

My first reaction was : Could Ruby store the floats like it stores the 
integers ?

I then read a bit of the ruby source code and besides reducing the size 
of the fixnum intergers, it does not seem possible.

NArray will probably provide the solution to most of my floating point 
performance needs.

Thanks again,

js