On Oct 14, 2006, at 8:41 PM, Devin Mullins wrote:

> Daniel N wrote:
> > I'm not sure that the thread ended up reaching a conclusion though.
>
> No, it didn't. The hypothesis was that the *parser* has a little  
> easier time with non-interpolated strings, and the benchmarks  
> didn't test that. They had N.times { "blah blah" } when they should  
> have had N.times { eval '"blah blah"' }.

We did that too:

http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/216376

James Edward Gray II