Long integers, xrange and number theory

[Martin v. Löwis]

Erik Max Francis <max at alcyone.com> writes:

> Don't you have these reversed?  If you've got long arguments, chances
> are building the actual list will be _huge_, so using range would be
> impractical.  

That would depend on the arguments, right? Changing range does not
require changes to the return type of the result; changing xrange does.

> xrange would be the different story, since it involves a proxy
> object that supports a sequence protocol.

Indeed. That proxy object would need to change, making it inefficient
even for small integers.

Regards,
Martin