@[email protected] to Programmer [email protected] • 4 months agoLike programming in bashlemmy.mlimagemessage-square185fedilinkarrow-up11.68Kcross-posted to: [email protected]
arrow-up11.68KimageLike programming in bashlemmy.ml@[email protected] to Programmer [email protected] • 4 months agomessage-square185fedilinkcross-posted to: [email protected]
minus-square@[email protected]linkfedilinkEnglish51•4 months agoI don’t normally say this, but the AI tools I’ve used to help me write bash were pretty much spot on.
minus-square@[email protected]linkfedilinkEnglish6•4 months agoFor building a quick template that I can tweak to my needs, it works really well. I just don’t find it to be an intuitive scripting language.
minus-square@[email protected]linkfedilink17•edit-24 months agoYeah, an LLM can quickly parrot some basic boilerplate that’s showed up in its training data a hundred times.
minus-square@[email protected]linkfedilink1•4 months agoIf When the script gets too complicated, AI could also convert it to Python. I tried it once at least, and it did a pretty good job, although I had to tell it to use some dedicated libraries instead of calling programs with subprocess.
minus-square@[email protected]linkfedilink25•4 months agoYes, with respect to the grey bearded uncles and aunties; as someone who never “learned” bash, in 2025 I’m letting a LLM do the bashing for me.
minus-square@[email protected]linkfedilink41•4 months agoUntil the magic incantations you don’t bother to understand don’t actually do what you think they’re doing.
minus-square@[email protected]linkfedilink14•4 months agoI wonder if there’s a chance of getting rm -rf /* or zip bombs. Those are definitely in the training data at least.
minus-square@[email protected]linkfedilink3•4 months agoThe classic rm -rf $ENV/home where $ENV can be empty or contain spaces is definitely going to hit someone one day
minus-square@[email protected]linkfedilink41•4 months agoSounds like a problem for future me. That guy hates me lol
minus-square@[email protected]linkfedilink11•edit-24 months agoIn fairness, this also happens to me when I write the bash script myself 😂
minus-square@[email protected]linkfedilink5•4 months agoYes, I have never wrote a piece of code that didn’t do what I thought it would before LLMs, no sir.
I don’t normally say this, but the AI tools I’ve used to help me write bash were pretty much spot on.
For building a quick template that I can tweak to my needs, it works really well. I just don’t find it to be an intuitive scripting language.
Yeah, an LLM can quickly parrot some basic boilerplate that’s showed up in its training data a hundred times.
IfWhen the script gets too complicated, AI could also convert it to Python.I tried it once at least, and it did a pretty good job, although I had to tell it to use some dedicated libraries instead of calling programs with subprocess.
Yes, with respect to the grey bearded uncles and aunties; as someone who never “learned” bash, in 2025 I’m letting a LLM do the bashing for me.
Until the magic incantations you don’t bother to understand don’t actually do what you think they’re doing.
I wonder if there’s a chance of getting
rm -rf /*
or zip bombs. Those are definitely in the training data at least.The classic
rm -rf $ENV/home
where$ENV
can be empty or contain spaces is definitely going to hit someone one daySounds like a problem for future me. That guy hates me lol
Yeah fuck that guy
In fairness, this also happens to me when I write the bash script myself 😂
Yes, I have never wrote a piece of code that didn’t do what I thought it would before LLMs, no sir.