3 ms·
Synthetic benchmarks are rarely useful. If you want to benchmark two languages you need to do this: Find N coders skilled in language X and N coders equally s
by InclinedPlane 15y ago
Synthetic benchmarks are rarely useful.
If you want to benchmark two languages you need to do this:
Find N coders skilled in language X and N coders equally skilled in language Y. Have all of them code several small but realistic systems of various kinds, then benchmark the results and compare. Better yet, benchmark real, live systems of comparable purpose built using different languages. Anything else leaves you open to having your results dominated by differences in level of coding skill, quirks specific to your synthetic benchmark, and effects that are only prominent for tiny, overly simplistic programs.