My question is whether it is good practice to include a unique wrapper phrase for custom commands and aliases.
For example, lets say I use the following command frequently:
apt update && apt upgrade -y && flatpak update
I want to save time by shortening this command. I want to alias it to the following command:
update
And lets say I also make up a command that calls a bash script to scrub all of of my zfs and btrfs pools:
scrub
Lets say I add 100 other aliases. Maybe I am overthinking it, but I feel there should be some easy way to distinguish these from native Unix commands. I feel there should be some abstraction layer.
My question is whether converting these commands into arguments behind a wrapper command is worth it.
For example, lets say my initials are “RK”. The above commands would become:
rk update
rk scrub
Then I could even create the following to list all of my subcommands and their uses:
rk --help
I would have no custom commands that exist outside of rk
, so I add to total of one executable to my system.
I feel like this is the “cleaner” approach, but what do you think? Is this an antipattern? Is is just extra work?
You have taken the first step towards creating your own distro.
Seriously though, what you suggest is fairly common but really a matter of preference. The same answer applies to “is it just extra work”.
I tend not to customize heavily because it keeps “me” generic and I can sit down at anything and be equally effective. Others heavily customize their environments to keep themselves productive and happy on the machines they actually use.
One advantage of your approach is you can create a “standard” user space across multiple distos. You do not have to remember if this system or that is Debian or Arch if “rk update” works everywhere ( even if is doing something different under the hood. This could be useful if you run a bunch of VMs or containers.
Do you have a favourite text editor that you heavily customize or do you use whatever? Same question for your DE. It is all scratching the same itch.
an easier approach to start with would just be to namespace them all with your initials when you set them as aliases, like
rk-update
,rk-scrub
— then you could tab-complete them instead of doingrk --help
. way less to maintain (unless you’re adding aliases from a bunch of different sources, in which case you may have bigger problems)What you described is a common practice to make oneself life easier.
If you want to write something like
rk
you would need to create a bash script and place it in your PATH so you can access it from there. It is fairly easy to do so, and you can back it up in GIT so that you have the latest version of the command line utility there. Even for the alias, I would say back it up in GIT, because you might lose them.Seems like a matter of preference, and I see the logic in it. I’ll mention that Nushell makes it easy to create custom shell functions that are invoked as sub-commands in this manner. https://www.nushell.sh/book/custom_commands.html#command-names
Good idea I’d say.
I have about 25 or so shell scripts I use somewhat regularly and well over 300 aliases. I actually specifically don’t wrap package manager related scripts for no reason in particular, but many often do.
My rule for an alias is if the amount of custom flags gets lengthy, and I use it often, yeah it gets an alias. Here’s an example of using yt-dlp:
alias ytdl='yt-dlp --sponsorblock-remove all --write-auto-sub -f "bestvideo\*+bestaudio[ext=m4a]/best" -f mp4'
For shell scripts, my rule goes that it should probably have multiple features related around a single idea, that way you can use getopts to create custom flags. For example, I have a script that wraps very basic, but commonly used,
git
commands, chaining the classicadd
,commit -m
, andpush
behind a series ofread
prompts, it has-h
flags for help-l
for a minimal log output,-i
to initialize a new repository (even using github api token to remotely create the repo if you want to use github), and-r
to revert back changes to a specified commit.Generally speaking aliases will get you what you need most of the time in a pinch, but shell scripting is more powerful, versatile, but potentially more time consuming.
Others have rightly pointed out that these abstractions can sometimes negatively impact muscle memory, but IMHO this only really applies if you work as devops or sysadmin, where you are often responsible for running many different Linux servers, but usually this isn’t an issue if you have access to the internet and can see your saved aliases and/or scripts (but yeah, instant recall of native commands trumps notes every time).
Additionally, another mentioned using
git
to keep track of your aliases, which I totally agree with. Whatever you do, back up your aliases and shell scripts, ideally with a git repo of some kind. This not only allows you to take your new scripts/aliases with you wherever you go, but also reference them later in case it’s not possible to use them on not your machine.Hope this helps. Bash can be crazy powerful of you take the time to learn it, and aliases are a great entry point to recognizing that potential. Here’s one of my favorites that combines
mkdir
withcd
:alias mkcd='{ IFS= read -r d && mkdir "$d" && cd "$d"; } <<<'
Good luck, and have fun.
Others have rightly pointed out that these abstractions can sometimes negatively impact muscle memory, but IMHO this only really applies if you work as devops or sysadmin
I feel both seen and called out lol
Lol. Emphasis on the H in IMHO. I meant no offense, it’s a valid point.
No offense taken, I just wasn’t expecting to get read like that lmao — I guess having to operate Other People’s Servers really causes a certain kind of (recognizable) cynicism
Disadvantages are, that you probably won’t learn the actual command and forget how to do it manually when needed. And that the Bash history will log the name of the alias instead full command.
I definitely think its a good idea to have a simple command running multiple commands, down to single letter changes as in
alias vim='nvim'
. Updating the system in particular needs an alias to me, because I combine much more with it (yay, flatpak, rustup) and at the end let balooctl6 check for new files to index:alias update='eos-update --yay \ && flatpak uninstall --unused \ && flatpak update \ && rustup update \ && balooctl6 check'
In general, I setup a shortcut to expand aliases to their target with CTRL+Space. That means if I type
update
and hit CTRL+Space, then it will be expanded to the entire list of what it would execute. This allows me to check what the command does and change something before execution. Also the history will log the long format this way, instead the name of the alias. I believe its this line in my .bashrc:# Expand alias with key binding "Control+Space". bind '"\C- ":alias-expand-line'
Sometimes I also use functions. Functions can’t be expanded. A distinguishing factor between Functions or Aliases set directly in your Bash is, that they have access to all variables and states in the .bashrc. A dedicated script would be much better isolated from your .bashrc file.
What I forgot in my previous comment: I have bunch of *rc aliases to open a specific config file with my neovim. It does not matter how the configuration file is named, with an extension, with rc in name, in what directory, it always follows the pattern singlecharacter+rc. Here some examples:
alias brc='nvim ~/.bashrc && source ~/.bashrc' alias nrc='nvim ~/.config/nvim'
As you see, it can even be a directory and not a file in case of nvim config. It would then show a filemanager on that directory with many config files.
Edit: Another thing I forgot (man I’m getting old) is that you can add a backslash at the front of a command, to run the actual command and not an alias. A common example is to have
alias='grep --color=auto'
andalias='ls --color=auto'
. If this gets in your way, just run the command as\grep
or\ls
to run the original command instead.You may like this pattern of starting all custom commands with a comma - benefits against a wrapper command would be shorter command names and built-in tab completion.
Personally I had to come to terms with the feeling that anything other than just running the raw commands will get me into trouble. I work on a lot of servers, and so I need to be able to rely on my shell knowledge even when my bashrc isn’t handy. So for me it became more about just remembering what software does what thing broadly, and then checking
man
for the finer details.But for a single personal machine, script it however you want. Just be aware that you’ll start to build muscle memory for aliases and custom functions that won’t follow you to new machines.