@b0rk_reruns ooooo these are super useful!

Another bash history shortcut I use all the time is !:s/PATTERN/SUBSTITUTION/, which lets you do sed-style regex substitutions on your last run command. You can also use ^PATTERN^SUBSTITUTION as a shorthand, or add the first few characters of the command you're looking for after the ! if you want to search for the most recent use of a specific command (i.e. !cat:s/PATTERN/SUB/)

There are a whole bunch of useful history shortcuts I didn't know existed until recently; definitely recommend looking up an exhaustive list if you're curious. This is the first time I've seen the ctrl+r one though, definitely gonna be using that frequently

@b0rk_reruns do NOT do loops like this! this is DANGEROUS

1. iterating over a list of files is fine. and if you know that your filenames don’t contain spaces, your approach will work. otherwise, you will iterate over a part of the filename, not the full one. Best case, your script doesn’t do anything. worst case, it spews garbage in your $PWD and you need to manually clean it up

correct way to do it:

IFS=$'\n'
for i in *.png; do
   convert -- "$i" "${i/.png/.jpg}"
done

IFS (inner-field separator) is a special variable which tells bash when to split stuff. By default, it’s a space; We set it to newline, because it’s a safe assumption that a file won’t have one.

1. try to prefix your file parameters with a double dash (--). this is a somewhat-standard way of terminating named arguments and specifying that everything after this point is raw data.

Terminating arguments is important if you use a variables that you don’t necessarily control. Take this for example:

a="-nastyFilename.jpg"
convert "$a" output.png

This will fail, because imagick will try to interpret the first arg as a switch. This can wreck havoc on your system, don’t ask how i know.

1.

always quote your variables, unless you truly want them to expand! if you don’t, they may get passed as separate args to a command you’re executing.

if your directory is HUGE (over 30k files), don’t use for; this is because for every command there’s an argument amount limit, and * expands to one arg per file. instead, do this:

while read i; do
    echo "$i"
done <<< $(ls)

another, even better way is to use find with its -exec option:

find . -maxdepth 1 -exec echo {} \;

anyways, sorry for the rant. i know a little bit too much bash, and i wish this curse upon everyone else >:3

@b0rk_reruns Also, having these:

## arrow up
"\e[A":history-search-backward
## arrow down
"\e[B":history-search-forward

In your `~/.inputrc` file makes it very easy to use history search on the current input and it's very intuitive. I will write `git` and hit the up and down arrows to navigate the git command history, for instance.

https://askubuntu.com/questions/59846/bash-history-search-partial-up-arrow