Duane Johnson pointed me to a very interesting discussion on Y Combinator about the differences between Scheme and Ruby. This is an excellent discussion--not a flame war--that I found enlightening. The summary, if you don't want to read the discussion thread:

• Ruby closures are more complex
• Macros, macros, macros...

Can't say enough about macros. Every language besides Lisp and it's close relatives trade macros for complex syntax. Maybe that's a good trade-off, maybe not. Nevertheless, it is a trade-off. You can't have the full power of macros without simple (abstract) syntax that's exposed to the programmer.

Now, that I've said that, I wonder if IO's impressive reflection can do the same. Someone more enlightened than I tell me: what can't be done with reflection that can be done with macros?

I do believe that using reflection is more complex than using macros. It reminds me of poking a stick in a black box to flip a switch.