Absolutely agree - was just being a little facetious. Clearly not about to start uglifying everything running on the server just because of a random benchmark o

Absolutely agree - was just being a little facetious. Clearly not about to start uglifying everything running on the server just because of a random benchmark of unrepresentative functions running in a loop. Correct to point this out, especially considering the predilection of devs to micro-optimise based on the most spurious of hypotheses. Putting a single IO operation in that loop will most likely completely change the picture, anyway - although have not run the test to check.