help-gnu-emacs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: not good proposal: "C-z <letter>" reserved for users


From: Jean Louis
Subject: Re: not good proposal: "C-z <letter>" reserved for users
Date: Fri, 12 Feb 2021 23:26:26 +0300
User-agent: Mutt/2.0 (3d08634) (2020-11-07)

* Dmitry Gutov <dgutov@yandex.ru> [2021-02-12 15:55]:
> On 12.02.2021 07:42, Robert Thorpe wrote:
> 
> > I agree with Jean Louis on that too.  I think C-z C-z is not good
> > enough.  Every other terminal application uses C-z by itself, it's a
> > convention.  It's been that way for decades.  If you write a terminal
> > application and do nothing special then C-z will suspend it.  That's
> > because it sends SIGTSTP.
> 
> Could we someday stop considering Emacs a "terminal application"? Yes, it
> has a version that works in the terminal, but it's limited in features
> compared to the graphical one.

Millions of users use computers and maintain systems on remote VPSes
and dedicated servers. Vultru, Gandi, Digitalocean, Servetheworld.net
and many other providers offer such systems. They are managed
remotely.

Remote computers are accessed by using terminal emulators. While
remote X program can run, that is not common to do over slower
Internet networks. I can easily run graphical Emacs from a remote
computer in local area network but cannot do the same from various VPS
providers, and why should I as such dedicated servers and VPSes are
meant to run servers and not X Window environments.

Those editors useful for system administration within GNU/Linux and
BSD derivative operating systems must always run on console or
through terminal emulators.

I cannot see any significant limitations of features. 

> 'C-x C-z', a binding which has existed for a long time, is not too hard to
> type.

Maybe it did exist, I have never used C-x C-z for shell job control as
shell job control is not perceived in Emacs in the first place, rather
in the description of various shells.

Korn shell job control uses Ctrl-Z:
https://www.ibm.com/support/knowledgecenter/en/ssw_aix_72/osmanagement/korn_shell_job_ctrl.html

Bash job control uses Ctrl-Z:
https://www.gnu.org/software/bash/manual/html_node/Job-Control-Basics.html

zsh job control uses Ctrl-Z:
https://gist.github.com/CMCDragonkai/6084a504b6a7fee270670fc8f5887eb4

There is POSIX standard on job control:
https://en.wikipedia.org/wiki/Job_control_(Unix)#Commands

"A job running in the foreground can be stopped by typing the suspend
character (Ctrl-Z). This sends the "terminal stop" signal (SIGTSTP) to
the process group. By default, SIGTSTP causes processes receiving it
to stop, and control is returned to the shell. However, a process can
register a signal handler for or ignore SIGTSTP. A process can also be
paused with the "stop" signal (SIGSTOP), which cannot be caught or
ignored. "

So Emacs could decide to ignore the shell job control and become
incompatible to shell users' expectations that Control-Z should work
when invoked from shell. The command is expected from shell, not from
Emacs. Emacs either complies, like it does not, or developers could
decide not to comply to those expectations.

That would be detrimental.

Jean



reply via email to

[Prev in Thread] Current Thread [Next in Thread]