If you're at the top of the buffer, guG will lowercase the whole thing.

So if you open a file, go to type G to jump to a line, but accidentally hit g, then try to undo it with u out of habit, before hitting G again, you do the same thing.

Uh to be perfectly honest for the past … 30 years I’ve ran with a buddies vimrc and really never took the time to understand it; it was perfectly good lol and to this day I could not recreate it if I lost it.

> lowercase a whole buffer,

Happens a lot to me actually!

That and accidentally incrementing a numeric value haha..

  > " Incrementally match the search.  I orignally hated this
  > " but someone forced me to live with it for a while and told
  > " me that I would grow to love it after getting used to it...
  > " turns out he was right :)
  > set incsearch

While I had times discovering myself to accidentally putting 'i' accidentally into MS Word docs and having vim key binding ls in VS Code (and formally eclipse) as well for that reason, I still feel clumsy. I think this is because I mostly rely on muscle memory rather than understanding deeply what I am doing to also allow me to go beyond what I normally do ) which already makes me faster than in any other editor). I found VimSpeak [1] interesting because I somehow understood how to really compose vim commands in a way

[1] http://www.youtube.com/watch?v=TEBMlXRjhZY

I miss keys sometimes, but if it makes bad edits (it doesn't always!), I just hit `u` to undo.

> that's a wrong mindset

The very idea there is a 'right mindset' is weird to me.

Nope! "G" moves the cursor to the end of the file. Very useful. Inferior editors have ctrl-end or alt-end, but with Vim, 90% of your lazy fingers stay on the home row!

Yes, it's exaggeration. Modal editing cannot read your mind.

I can see how that could work depends on the setup and the context. For example: People might use `. to jump to the last edit, or to a mark they set manually. Or simply `ciq` to edit inside the next quote without any manual cursor movements. I see people use plugin like harpoon to jump to their favorite location quickly. If you don't know about such setup, seeing people type <leader>1 to jump not just within a file, but across files, seems magical.

>> single keystroke could move the cursor halfway across the file to exactly the right spot.

> sorry what is this? exaggeration?

It is indeed. I need two keystrokes to move from an arbitrary positions in an arbitrarily long file to the exact spot I need to be.

I use it all the time, in fact. Multiple times an hour. It's muscle memory now for me, while reading/navigating code, to automatically do `ma` or `mb` etc.

At some later point I realise "let me read the function definition again" and then I do `'a` or `'b`, etc.

Although, also, perfection probably doesn’t come from adding options either.

I can use hjkl y and d perfectly! :set rnu and I can even throw some numbers in there!

The problem is both (1) knowing what you want, and (2) specifying what you want.

(1) is hard enough and a necessary prerequisite to (2) which, even so, is even harder than (1).

Good, documented software is the accumulated knowledge of people who (1) knew what they wanted, (2) implemented it, and (3) communicated how it works. AI can ease the building of such software but does not make the process trivial.

Learning Vim has been one of the highest-longevity skill I picked up in University. With every new technology - autocomplete, IDEs, various GUI design interfaces - there's always a chorus of folks who say: "Well now you don't need to Vim, this new tool makes that obsolete." And every time I end up having to manipulate a mountain of text regardless - whether that's in logs, source code, configuration, or documentation. With the amount of text that AI outputs I don't see the need to manipulate text going away and Vim is one of the fastest and most flexible ways to do that.