On Wed, Oct 21, 2009 at 5:16 PM, Roger Pack <rogerdpack / gmail.com> wrote:
> Yeah I think that's the case in 1.9, as well--if they ripped out the
> fiber based enumeration it would be much faster. Wonder if there's a
> way to get the best of both worlds...

The only way I can think of would be to have a stackless
implementation, like both JRuby and Rubinius have had in the past. But
in both cases, the performance hit from being stackless greatly
outweighed the benefit.

To be honest, Python may have it right here, since you can only yield
from within the first level of a coroutine, which means that
implementations like Jython can just make that one level of logic
resumable. In JRuby, because the Generator/Fiber can resume at any
level of logic, we would have to make the entire system be stackless
or resumable, all the way through.

- Charlie