Comments for vvs's jruby  

Other than maybe generating slightly smaller bytecode, I don’t really see how this could help. A tiny, final method should be the first thing the JVM would inline, resulting in the same code in the end (save a bit of stack juggling, which would disappear when everything is allocated to registers anyway).

I haven’t benchmarked or anything, though; this is merely from observing the changeset, as I can’t do much more than that this week. Is there something I’m not seeing here?