zlacker

[return to "The Case for Nushell (2023)"]
1. BadBad+Wo[view] [source] 2026-01-07 17:48:36
>>raveni+(OP)
I have always struggled a bit with shell replacements. I use zsh but only because of oh-my-zsh. Apart from that I always thought of shell scripts as a necessary evil for interoperability. Today you can usually expect a linux system to have bash or at least sh. That is why I do shell scripting. Nushell is usally not preinstalled on a system and so I cannot expect it.

If I want to do real scripting/programming I use python or another dedicated programming language. I don't really know what the value of Nushell is for me. Maybe the plugin system is amazing but at the moment I miss nothing in my zsh.

◧◩
2. timeon+GD[view] [source] 2026-01-07 18:40:31
>>BadBad+Wo
It is not just about scripting but the structure of input and output. Also no need to give up (z)sh switching shells is eazy.
◧◩◪
3. BadBad+xF[view] [source] 2026-01-07 18:49:19
>>timeon+GD
Maybe I need to get into nushell more but my feeling is that if it's a simple problem it the benefit will be negligible and if it is too complicated for that I will probably write a python script.
◧◩◪◨
4. xpe+aq1[view] [source] 2026-01-07 21:56:28
>>BadBad+xF
That is how I felt with Zsh, even though Zsh scripting is a step up from Bash. But with Nushell there is a huge space of scripts I now prefer to write in it.
◧◩◪◨⬒
5. BadBad+kJ2[view] [source] 2026-01-08 08:20:03
>>xpe+aq1
I use zsh because of oh-my-zsh. I don't any of the scripting features.
◧◩◪◨⬒⬓
6. xpe+b17[view] [source] 2026-01-09 15:50:22
>>BadBad+kJ2
Please take this with a dose of humor (with which it is intended) ... I wonder if part of you prefers to not see Nushell for yourself? This is totally understandable; in the face of too much information, cutting off branches of investigation can be a good coping strategy.

I freely admit this is how I felt about – to talk about a different technology – the Zig language for at least a year. Until one day I spent about an hour trying it out. I am not a "convert", but it yielded useful personalized insight.

Speaking personally for a moment, when I reflect on my higher-level philosophy, it tells me: "If you notice yourself only «generating a list of reasons why NewThing doesn't fit your existing needs», you might be in a cognitive rut. Current needs matter, but so does ongoing capability development. And psychologically one's awareness of tools tends to shape what we perceive as needs.§"

So, it continues, "Instead, do a timebox and try the thing.* Attempt to see it for what it is rather than relative to what you already know! Doing this is really hard, so you'll probably need to seek out techniques to get better at it over time."

* Assuming it doesn't involve drugs, chemicals, or software supply chains of unknown origin.

§ When you only have a hammer...

[go to top]