Battle of the Elixir code editors

Atom

####Right:

• Well supported with plugins and themes
• Can provide something very IDE-like
• Doesn’t take additional brain-power to use (keyboard shortcuts)

####Wrong:

• Memory hog
• Splits are really unintuitive having used tmux/vim/spacemacs/etc

Neovim

Right:

• Modal editing, once you’ve used it, you’ll feel crippled anywhere you can’t use it
• Plugins: can find (or make) a plugin for almost anything
• Non-blocking actions with Neovim
• All the other Vim goodness…
• Fast to start (I’m looking at you Atom & Spacemacs)

Wrong:

• Plugins: confusing to configure when you’re starting out, or trying something new
• Buggers up Phoenix’ live reload., due to Neomake. Live reload is more important to my workflow, although Neomake was a revelation when I first added it.
• Extra brain cycles required whilst you’re learning either Vim or the project at hand. I still find Sublime/Atom easier to hop around and get to know a codebase

I’ve used Spacemacs, Sublime, Webstorm (et all), VS Code. Didn’t really like any of them. Special mention for Coda though - anyone remember how cool that was?

I’ve heard that argument over the years many, many times, and am still waiting for it to happen

META NODE: A quick comment which turned into many paragraphs. I am getting old

The easiest way to ensure that you will, is to constantly
use Vim keyboard shortcuts.

I attribute vim keyboard shortcuts for saving me from any carpal tunnel
syndrome or other problems. Before I went full vim on everything (i3wm,
mutt, vimb browser, vim, vi-mode in shell) I had much more problems with
muscles in arm and back, reaching for the mouse all the time instead of
keeping my hands on the home row where they are supposed to be. Nowadays
my body like me better.

Also, I try to rely on tools which hopefully wont change to much. I was
going to say that “This lets me focus on developing rather than adapting
to new tools” but I don’t actually focus on development anyway so it is
not a good point At least I don’t have to spend energy to change my
tools for no added benefit.

I think everyone wants development to happen in their own particular
field (me included) but that overall most people do not want drastic
change. Looking at the success of Windows XP for example I think most
people still would have preferred to use their old tools. And to be
frank I don’t actually think end user software has actually improved
much in the past 20 years. I think software has become victim to the
markets in that you need to have a new shiny thing to be able to sell
it. The easist thing to change to make something look better is to
change its apperance.

20 years ago I used emacs, gcc, vim, Word, Excel, Netscape or
IE. If you compare them with today not much has actually changed I can
easily do whatever I do today with yesterday tools. The extra hardware
we have today is eaten up by sluggish software. The one thing that has
changed is improved internet speeds and the ability to use “cloud”
storage and the like.

I’ve used the same setup for over 10 years and I hardly have to change a
thing. This let me master my tools. Who cares if learning curve is hard
if you invest over such a long time. The important thing is that you
actually can become a master of the tools. Most things today are
developed for beginners and that is where things stop. The tools wont
allow you to become better, it is all sacrifised to be beginner
friendly.

To tie this back into editors: I use vim