The problem with that attitude is that real-world usage is what matters if you're benchmarking or evaluating for your own use. The vast majority of people who

The problem with that attitude is that real-world usage is what matters if you're benchmarking or evaluating for your own use. The vast majority of people who use Ruby for web development are using Rails (because Rails is a large part of what makes Ruby effective and attractive for web development - in general, YMMV), so you can't make realistic performance comparisons without comparing commonly-used stacks. Otherwise it's just a useless pissing contest.