Re: Concern about new binding.

[Juri Linkov]

> I like `C-x r e' -- it has pretty good mnemonics, and it's not too
> cumbersome to type.
>
> I also like the idea of adding a new prefix command for buffer-related
> commands.  The list of proposed commands we could bind there was pretty
> compelling -- it had a whole bunch of commands that (I think) people use
> somewhat frequently, and grouping them that way will help with
> discovery, I think.
>
> `C-x x' seems to be in the running here, and is pretty
> convenient to type.  If we go with `C-x x', then I guess `C-x x g' would
> be the binding for `revert-buffer', I guess?  (To mimic the `g' in
> special-mode-map.)

I don't know, what was the reason for freeing `C-x x' some time ago?
In etc/NEWS.22:

  ** The register compatibility key bindings (deprecated since Emacs 19)
  have been removed:
    C-x /   point-to-register (Use: C-x r SPC)
    C-x j   jump-to-register  (Use: C-x r j)
    C-x x   copy-to-register  (Use: C-x r s)
    C-x g   insert-register   (Use: C-x r i)

What if some package becomes popular in the future and wants to claim
the `C-x x' prefix map that has good mnemonics for it?  Such as
xwidgets.

In no one has such plans in the foreseeable future, then it would be
reasonable to use the new prefix map to bind as many keys as possible.
There are many frequently used commands that are still unbound.  Then
we could group the keys by categories, e.g.:

Buffer-related commands:
C-x x b r - revert-buffer
C-x x b R - revert-buffer-with-fine-grain
C-x x b n - rename-buffer
C-x x b u - rename-uniquely
C-x x b c - clone-buffer
C-x x b i - insert-buffer

File-related commands:
C-x x f a - append-to-file
...

Window-related commands:
C-x x w p - previous-window-any-frame
...