Re: Concern about new binding.

[Kévin Le Gouguec]

Lars Ingebrigtsen <larsi@gnus.org> writes:

> The one concern about the `C-x g' binding is that Magit already
> recommends it, but it's unclear to me how many people actually use it,
> and what it's bound to.  Is it just a global binding for `M-x magit'?
>
> Presumably Magit users who've bound it to that will continue to do so...
> and then they'll miss the new binding(s) under `C-x g', but I guess
> that's up to each individual user.

To clarify:

- C-x g is bound to magit-status, which is Magit's main entry point,

- Magit includes an autoloaded form that binds C-x g if
    - that key sequence is not bound to anything else, and
    - magit-status is not already bound, and
    - the user hasn't set an explicit "dont-do-that" variable.

(Same goes for two other bindings: C-x M-g for magit-dispatch, and C-c
M-g for magit-file-dispatch.)

So adding a default binding for C-x g *will* change how Magit behaves in
its default configuration.


I struggle to form a solid stance about the change under discussion:

- I wouldn't find it outlandish for Magit to do something similar to
  rg.el: provide a function (say magit-enable-default-bindings) that
  users can call in their init file to easily setup some bindings under
  a prefix (that would default to C-c g).

- I wouldn't mind C-x g (or C-x g g, or C-x g r) being bound to
  revert-buffer.

- I find C-x g somewhat awkward as a prefix for buffer commands.  Not
  really mnemonic, at least.