Oi

Lists
mandatory
Write a script that lists all files (even ones with names beginning with a period character, which are normally hidden) in the current directory and the parent of the working directory and the /boot directory (in this order), in long format.

Be careful with the /

Repo:

GitHub repository: holbertonschool-shell
Directory: basics
File: 11-lists

requirements:
Allowed editors: vi, vim, emacs
All your scripts will be tested on Ubuntu 22.04 LTS
All your scripts should be exactly two lines long ($ wc -l file should print 2)
All your files should end with a new line (why?)
The first line of all your files should be exactly #!/bin/bash
A README.md file at the root of the repo, containing a description of the repository
A README.md file, at the root of the folder of this project, describing what each script is doing
You are not allowed to use backticks, &&, || or ;
All your scripts must be executable. To make your file executable, use the chmod command: chmod u+x file. Later, we’ll learn more about how to utilize this command.
and the error it shows:
[valid_number_of_lines] Desired number of lines: 2
[valid_number_of_lines] Actual number of lines: 4
[valid_number_of_lines] Last 5 characters of the file: "done\n"

Hi

I had a different experience before. I always used ChatGPT, Gemini Advanced, and LLaMA 3.1 400B. They rarely provided code with minimal errors, expecielly in complex code. With ChatGPT, I often had to edit the code, and sometimes it would repeat the same code that had errors, even though I had already told it about the error along with the error message. But it still gave me the same code again. With Gemini, it was worse; it refused to provide code and responded with "I’m just a text model." And like GPT, it rarely gave error-free code. There were even times when it gave unrelated code together, for example, the first four lines of code were about A, but then the following lines were about something unrelated to the code above. With LLaMA 3.1 400B, it was the same as the others—the code had errors. But when I tried Claude Sonet 3.5, bruh, I was moved. It could understand complex code and didn’t have errors