Most developers know that curl has more features than wget, but for any task involving downloading files, wget seems to be the most popular. Why is that?
Well, it really boils down to the defaults. While the author of curl has
his design choice, curl defaults to printing to stdout whereas wget defaults to
saving to a file. Naturally, people will take the path of least resistance, and
I bring this example up because, despite the intentions of the authors, users tend to use the tools in whatever way works best for them.
According to the man page, touch's main purpose is to change file timestamps. When was the last time you used it that way? I can confidently say in my decade and a half of using unix tools, I have only had to do that once (and it was for a shell programming homework assignment).
Under "description" in the man page, it says:
Update the access and modification times of each FILE to the current time.
A FILE argument that does not exist is created empty, unless -c or -h is supplied.
Aha! Interesting that what seems like just a side effect (creating an empty file) has become the most common use case.
It is conventional wisdom that cat is used to print out files to stdout because cat stands for…concatenate? Huh?
Imagine you were to write a function that takes in strings and returns the concatenation of all of them. A mindful developer would take into account any number of inputs, including zero and one input, but a beginner (or lazy) developer might not. From this problem description, it would seem like inputs of zero or one strings are edge cases.
Strange how for cat, concatenating one file is its primary use case!
The man page for cat says it will concatenate files and print on the standard
output. However, most people use it to simply print the contents of one file
to stdout, which has led to using cat to chain commands,
cat file | grep "hi".
This idiom is so common that people have even written about the Useless Use of Cat and the Useless Use of Cat Award!
Despite the title of this post, I think it's perfectly fine to use the tools in anyway you need. In fact, I would say these tools are the easiest to use for the functions I've mentioned above (which I use all the time). I just find it quite fascinating to see how in both touch and cat, the intended use has been neglected (or even forgotten) because of a more useful side effect.
In the case with cat, I find there are benefits with the "useless" use. Conceptually, chaining simple commands through pipes is much easier than remembering which argument is for the input file. Also, when building a more complicated command, I like to visually see the output step-by-step. And with the only con being…spawning an extra process? It's okay, I'll stick with my current workflow :)