RE: [External] : Re: PROPOSAL: Repurpose one key and reserve it for third-party packages

[Drew Adams]

> I realize I missed something... What has happened?
> Can anyone summarize it i one or to paragraphs
> instead of ~20?

Here's my summary (1-9).  Others might see it
differently.  And it's possible that I don't
remember some details perfectly.  Someone will
correct me if I'm mistaken somewhere.

1. Bug #46151, "Set revert-buffer-function in shell
   command output buffers", proposed binding a key for
   `revert-buffer' in shell command buffers (only).

https://debbugs.gnu.org/cgi/bugreport.cgi?bug=46151

2. A discussion in that bug thread led to discussion
   of whether Emacs needs a _global_ key binding for
   `revert-buffer'.

   My own opinion was this:

     Overall, my opinion is to NOT bind it by default.
     Yes, it's useful.

     But it's also easy to do with `M-x revert'.  (And
     then repeat that as a previous command, as needed.)

     And it has many existing bindings, for modes where
     it's used frequently.

   Some others agreed that we need no global binding
   for `revert-buffer'.  If interested, see the thread
   for the discussion.

3. The bug-thread discussion then went toward binding it
   to a global key under `C-x'.  My position was not only
   that we need no global binding for it, but that `C-x',
   in particular, should at this point be left alone.  I 
   explained that I already was forced to move from using 
   prefix key `C-x p' to `C-x x' etc.  I said:

     Users and 3rd-party libraries should really start to
     take precedence now, IMHO.  Emacs should try not to
     bind any more keys by default - starting with `C-x'.
     And certainly for things like `revert-buffer', which
     have survived for 35 years without a default binding.
     YAGNI.

4. Maintainer Lars bound `revert-buffer' globally to
   `C-x g', and closed the shell buffer bug.

5. User Ergus posted in emacs-devel about that,
   complaining that the question should have been
   discussed in emacs-devel (the consequences are
   wider than just shell buffers).

6. A long discussion ensued in emacs-devel.  In
   that discussion I agreed that when a bug-thread
   discussion moves beyond the bug to wider questions
   with wide consequences it should preferably be
   moved to emacs-devel.

   And I repeated my disagreement about globally
   binding `revert-buffer', and in particular about
   binding it to something on `C-x'.  I repeated my 
   suggestion that Emacs desist for a while from
   binding any new keys - at least that it try and
   have that as a convention/goal, and that we
   reserve those remaining keys for use by 3rd-party
   libraries.  Gregory proposed instead that we
   just reserve one key under `C-x', for use as a
   prefix key by 3rd-party code.

7. The main maintainer, Eli, disagreed that questions
   wider than a bug's subject should generally be
   brought to emacs-devel, and he supported the
   decision to bind `revert-buffer' to `C-x g'.

   Other users spoke up complaining about that key, 
   suggesting other keys for it, and so on.  Each
   time a key was suggested someone invariably
   complained.  (Not I - my disagreement is more
   general - I would no global key to be bound to
   it by default.)

8. There was also discussion about the problem of
   people in emacs-devel not being aware of the
   details of this or that bug report, IOW, _how_
   to make the wider group be aware of some bug
   discussion that's grown wider and should maybe
   be moved to emacs-devel.

9. That's pretty much where we are, I think.  Again,
   I may have forgotten something here or there.  I
   haven't intentionally left anything out.

___

Here's the summary of my position, from that bug thread:

https://debbugs.gnu.org/cgi/bugreport.cgi?bug=46151#88

And here's a longer post of mine about the various
questions - a summary, but not short.

https://lists.gnu.org/archive/html/emacs-devel/2021-02/msg00312.html