Stl Linux Unix Users Group

Imagine if the output from a command was structured data, e.g. like a SQL table, instead of plain text.

Fifty years ago, Unix implemented pipes, where the output of one command was fed as input to the next command. The data flowing between the commands tended to be plain text and often required parsing by intermediate programs in the pipeline. Furthermore, that text was often organized by lines and rows. And if it wasn't, that made parsing the text a bit of a challenge. Despite those limitations, the concept was revolutionary. But imagine if that data was more than text. Imagine if it was structured data so that you could easily query, select, filter, and sort the data regardless of its structure.

Enter nushell, a different way of piping data. Nushell also uses pipes, but the data flowing between commands is structured data, e.g. text, CSV, JSON, YAML, SQLite, Excel.

And you can use common nushell commands to operate and transform the data. It's like using a shell and a programming language from a command line.

In this talk, we will casually walk through using nushell with plenty of examples of what it can do. Nushell is still very much in the early phases of active development, but it shows a lot of potential for working with structured data.