> The author even points out "programs as pipelines" which is literally the UNIX philosophy. Yes, and if the thing I'm trying to do has a small input, it will

> The author even points out "programs as pipelines" which is literally the UNIX philosophy. Yes, and if the thing I'm trying to do has a small input, it will only be done once, etc. I will often just pipe `grep` to `sort` or whatever, because it's less typing, it's generally clearer to a wider range of people, etc. But on larger inputs, or even things like doing a single pattern inversion mixed with a pattern match, I like awk.