Re: Invoking Magit

[Karl Fogel]

On 02 Feb 2021, Stefan Monnier wrote:
> Those are for use *within* the major/minor mode.  The keybinding 
> under discussion is a global one to make it easier to enter 
> Magit.  It doesn't have much choice but to collide with something 
> :-( 

Yes, I know?  (Sorry; I think you maybe thought I was saying 
something I wasn't saying.)

> > > I think it'd make more sense to use a keybinding under the 
> > > `C-x v` prefix. 
> > ...I couldn't tell whether you meant it would make more sense 
> > for *users* to 
> >  do that (i.e., by their choice), or for *Emacs* to do that 
> >  (i.e., we 
> > reserve that slot for Magit). 
> Neither/both.  I meant for Magit to do that. 

What I'm proposing is cooperation between Emacs and popular 
independent 3rd-party packages:

Emacs reserves certain bindings for a given package, so that when 
the package maintainers use those bindings, they know it's safe. 
Emacs would maintain a registry so it's clear to everyone what's 
going on.

Let me put it this way:

If Magit were shipped as part of Emacs, there would be no question 
in about what to do, right?  Emacs would choose keybindings for 
the Magit entry points, the same way Emacs chooses keybindings for 
anything else.

All I'm saying is, the fact that Magit is not shipped with Emacs 
is largely irrelevant here.  What matters is its popularity, not 
its provenance.  Emacs can still reserve keybindings for it, and 
the fact that the package is supplied from elsewhere is fine.

(What those bindings would do when the package isn't loaded is 
simply display information about where to get the package in 
questions.)

Best regards,
-Karl