Re: Should `revert-buffer' preserve text-scaling by default?

[Karl Fogel]

On 02 Dec 2019, Stefan Monnier wrote:
>It will let people try it out, so we get experience.  We could also
>change its value tentatively in Emacs's `master` after we cut the
>emacs-27 branch, so we have time to hear complaints/bugreports in order
>to decide if we want to change it officially in Emacs-28.

Ah, that makes sense.  I didn't think of the possibility of changing it in 
'master' once the release branch is cut.  That answers the question I asked 
Eli: how does this two-step process help us?  You've explained how it does.

>There's still a UI question, tho.  There are (at least) 3 operations we
>want to bind to the same command:
>
>    A- revert buffer content
>    B- revert to auto-save file
>    C- revert and reset modes
>
>Currently C is the default and B is obtained via C-u.  With the bool var
>suggested, we could have A by default but would C-u give you C or B?
>In that case would we still give access to the other somehow?
>
>[ And of course, there's option D which is "revert to auto-save file bur
>  without resetting modes".  ]

Well, I was thinking we wouldn't change the current C-u behavior.  We'd just 
add a new defcustom, `revert-buffer-reset-modes', defaulting to `t' for now.  
If someone wants to change the mode-resetting behavior, they just change that 
variable.  This is not as easily interactive as a prefix arg would be, but I'm 
not sure we need such convenience here (unless we believe that people are 
likely to change their preference from call to call, but I suspect that's a 
rare case).

I'm not going into detail here about how `revert-buffer-reset-modes' would 
interact with the `preserve-modes' parameter; that implementation question is 
easy to resolve.  I'm just proposing the following user-visible behavior:

1) C-u continues to control revert-to-auto-save-file

2) New defcustom `revert-buffer-reset-modes', defaulting to `t'

Thoughts?

Best regards,
-Karl