Hacker News new | past | comments | ask | show | jobs | submit login

I had a similar realization early after first picking up vim in college: customization of any tool eventually hits a point of diminishing returns beyond which further alterations reduce your ability to use the tool in its default state. It's an insight I've found applies to almost any tool... From software to hardware and beyond.

Master the default behavior of a tool, and then improve your effectiveness with customization, but not so much customization that you can no longer use the tool effectively in its unmodified state. Sometimes you have to use other people's tools, and it's important to still work effectively when you do.




I agree with this very much! I learnt it the hard way: my experience with many powerful tools, Vim included, was to learn the basics first, customize it to the point where you can't recognize it anymore, and then finally strip it down to the basics again, keeping only the configuration that doesn't prevent me from using the the raw features.

I used to manage complex "dotfiles" and scripts to configure a new computer, etc. I still technically do, but they are much more simple now. I just don't want to spend my time on configuring the stuff anymore, and appreciate the out-of-the-box experience much more. This by itself became a criterion when choosing new tools, frameworks, etc.


I used to think like that, but 10 years ago I decided to create a git-repo for my .emacs.d and started configuring beyond the most trivial settings (and including all dependencies in the repo too). Diminishing returns? Not sure how to measure. Editor configuration was never anything I did because I thought it would somehow save time, as some seem to imply, but rather just about removing sharp edges and making the experience of editing files nicer. With everything a quick git clone away it is rare that I have to work without my configuration.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: