Here's the thing: these arguments all too commonly focus on subjective notions of "simplicity", and toy examples divorced from actual common practise, and or solid comparable benchmarks.
Show me a range of practical examples, for each competing env (awk, sed, perl, python, ruby, maybe bash).
Include:
- time it takes to teach a total novice (the time it takes to learn whatever is needed that example, not the entire language)
- how easy it is to recall said knowledge at a later date
- how fast example is multiplied by how much you are likely to use it = actual time saved in terms of execution. for small, fast examples the difference is irrelevant, a 10x speedup that is 0.1s vs 0.01s is meaningless.
- how extendable an example is. Hence the original example should include a series of extensions to the original task, to demonstrate how flexible / composable they are: e.g. task 1) count lines in a file; task 2) count lines in a file, then add 42 to it;
I suspect awk falls behind in practicality vs perl (which can do simple one-liners, but also more complicated constructs), but perhaps has a hidden virtue wrt speed in more expensive tasks, ala https://news.ycombinator.com/item?id=17135841 or https://news.ycombinator.com/item?id=20293579