I love the power of the vi/vim modal editing approach, and I applaud how
far Evil (and all prior efforts, especially viper) take it in Emacs.
I thought I'd write about why I'm not going to use Evil. I tried it for
about a week. In the past, I've used viper for about a year before
dropping it in favor of vim, which lasted another two years. Overall I
have used Emacs for 18 years.
In short: it is easier to use one editor at a time.
By default, Evil presents you vim key bindings, but makes you use Emacs key
bindings too:
(a) see the huge list of modes in evil-emacs-state-modes
(b) insert mode has Emacs key bindings
On (a), most of these modes don't completely rebind the Emacs key map. In
particular, they usually don't redefine how the point is moved, how
buffers, windows and frames are managed, etc -- this is all vanilla Emacs
stuff. In contrast, equivalent modes in vim use the vim key bindings and
ex commands for doing all this stuff.
On (b), Evil gives me M-q, C-a, C-v, etc. while in insert mode. This means
I'm using two editors at once: vim and Emacs. In Evil, many very basic key
bindings do wildly different things when switching from normal to insert
mode and back: C-f, C-b, C-v, etc. Vim doesn't suffer from this (much)
because insert mode is relatively free of key bindings.
The vi -vs- Emacs dichotomy is always there, creating friction. This is
certainly not the way Emacs would have designed its key commands were modal
editing its initial design goal.
I don't know what the solution is. In my ideal world, all Emacs key
bindings would be entirely blown away and replaced with vim-ish ones.
That's pretty audacious!
