> I think it's written by someone who finds pipes and awk too awkward to work with.

Actually it's exactly the opposite, it's born out of a love for pipes, and shells, and tools like awk. If you know anyone working at Amazon, ask them to search "11 years of work in a weekend" for a tale of shell heroics that I wrote about while I worked there.

dt is intended to be a new tool in the "shell one-liners" category, just with concatenative FP semantics. :) It will not be everyone's cup of tea, and I will still love and use awk when appropriate

> I think it's written by someone who finds pipes and awk too awkward to work with.

This one of substituting AWK/shell/sed/Perl with a forth-like lang is a good idea in a sense, it doesn't break the flow because presentation comes later and logic is at the beginning of the dt part, with the aforementioned tools you have the logic and output mixed all over the place. I will however still use AWK.

> ...it doesn't break the flow because presentation comes later and logic is at the beginning of the dt part

I didn't mean to discredit the work done. It's a big undertaking in any case. The idea and the aim is good, however it breaks the conciseness and reduces the speed of implementation.

> with the aforementioned tools you have the logic and output mixed all over the place.

I think this is a secondary effect of composability, pipe and conciseness requirement.

> I will however still use AWK.

Me too, and this is why I made my prior comment, exactly.