Re: Concern about new binding.

[Jean Louis]

* Stefan Kangas <stefankangas@gmail.com> [2021-02-06 18:24]:
> Gregory Heytings <gregory@heytings.org> writes:
> 
> > Yes, it is unavoidable that some people will be against changing a
> > binding.  I have no preference between the three proposed keys, and
> > anticipated that there would be more objections against using "t" for that
> > purpose.  If we put "t" aside, there are still two other options: "z" and
> > "o".
> 
> FWIW, I would not like to see C-o change as I use it daily.  But I could
> live with it as I can rebind it locally.  It would be too bad that we
> would then lose the nice symmetry between `C-o' and `C-x C-o'.
> 
> I'm in favour of rebinding C-z to exactly one thing: `undo'.  It would
> IMO be very unfortunate if we do rebind it yet miss out on the
> opportunity to make Emacs more like other applications.

For console users it is good to consider why people use C-z in the
first place.

Why would anybody wish to "stop the job" in the shell?

Maybe there is urgent need to replace disks, mount disks, unmount
such, connect to remove servers during job execution?

Maybe there is need to replace the video file in the file list while
not interrupting the processing of other 20 video files? Maybe such
video file processing takes 2 days. I am using Emacs Lisp to process
video files and this is real world example. Sometimes I process video
files on remote server. Maybe I do not want to process video files
during the day and job can remain suspended until the night without
interrupting the queue of video files being processed.

Thus it is good to consider the purpose for users to suspend a job in
their shell.

Yes, sure there are other Emacs key bindings to suspend the job, but
suspending a job should be compatible with shell bindings to suspend
the job and that is C-z.

Breaking incompatibility puts users' data at stake.

Jean