[plt-scheme] Re: Project Euler was Hash Table implementation

David Einstein wrote:
> I can't wait.  I happened across your original srfi-40 by accident
> after I had written sorrier versions of the extended functions.  I'd
> be interested in playing with whatever you have, and I am sure that
> there are others that would as well.
> 
> I was the one who said that srfi-40 streams are tricky.  The fact that
> they do not always delay when you expect them to [...]

It goes beyond being tricky.  The document and the reference 
implementation are inconsistent in a fundamental way.  For example, the 
document states:

    "no element of the stream is evaluated until it is accessed."

And yet, the following causes an error because the car element is 
evaluated even though it is never accessed:

    (stream-cdr (stream-cons (1) stream-null))

David