choose

This idea sounds like a good use for ‘choose’ [1]. It’d be a bash script that uses curl to get a list of the languages with gitignores, pipes to choose to let the user choose one, and pipes it into curl again to get the gitignore contents and sends to stdout. [1] https://github.com/sdegutis/choose

In a similar vein, I wrote choose[1] which I thought was very novel at the time, since it is a command line utility that pops up a native GUI (for fuzzy matching) and takes its input from stdin, lets the user choose an item by interacting with the native GUI, and spits output onto stdout. I thought it was the coolest idea since you can integrate it with any CLI that can launch apps and read/write to standard input/output, including vim and emacs. [1]: https://github.com/sdegutis/choose

I've written probably a hundred open source projects in my short professional career. But only two of them do I actually use every single day: a simple OS X window manager [1], and a hybrid command-line/native-GUI fuzzy string matcher [2] (you'd be surprised how many ways/places this can be used). The rest of them didn't serve well as practice like I'd hoped, they were mostly just a waste of time I could have spent better elsewhere. So lately I have new criteria for how and when to write open source projects: (1) when I realize that I have need to automate something, (2) and it won't take more than a weekend to get a basic functioning version up and running. Anything else, and I nope on out of there. I've got too little time and too many things to do already, I'm not about to waste it creating more unnecessary software. The world has enou