On Tue, Oct 25, 2011 at 4:27 PM, Joshua Ballanco <jballanc / gmail.com> wrote: > This was the subject matter of my RubyConf talk. You can view the video > here: http://confreaks.net/videos/671-rubyconf2011-keeping-ruby-reasonable . > If you skip to 19:33, I presented a very real example of how this property > of Proc#bindings can have an impact on performance. Awesome, thank you for the link Joshua! Hopefully this helps clear up that it's not a "JRuby can't optimize it so it's bad" kinda thing. It's actually a very real optimization headache for *any* implementation, including MRI itself. And then there's the encapsulation-breaking aspects, which are pretty nasty too. - Charlie