The issue with Unix pipelines is that they handle arrays of structured data very poorly. Since everything is just a stream of bytes, non-trivial pipelines often

The issue with Unix pipelines is that they handle arrays of structured data very poorly. Since everything is just a stream of bytes, non-trivial pipelines often need to shuffle data around to make the next stage happy because there is no standard (is it a tab, comma or space this time? newline or null byte?). For all its quirks, writing pipelines in PowerShell is usually far easier because they are a stream of objects and filters can directly work with property names, instead of fiddling with cut until everything is aligned properly. I'll admit I've never used awk so my user experience of Unix pipelines is probably a bit worse than it has to be.