Re: bin/58687: sh: implement suspend builtin

["Valery Ushakov via gnats" <gnats-admin%NetBSD.org@localhost>]

The following reply was made to PR bin/58687; it has been noted by GNATS.

From: Valery Ushakov <uwe%stderr.spb.ru@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: bin/58687: sh: implement suspend builtin
Date: Sun, 22 Sep 2024 23:57:23 +0300

 On Sun, Sep 22, 2024 at 20:00:03 +0000, Robert Elz via gnats wrote:
 
 > That is the kind of thing which user specific functions are
 > good for.
 
 Oh, I agree about functions.  But for "suspend" there's prior art that
 goes way back and it's in people's muscle memory.  And I usually run
 into this on systems where /bin/sh is not my shell, so I don't have
 any startup files.  Hence this bug to have this defined "by default"
 in any/interactive(?)  invocation of sh.
 
 Plus a builtin may add extra safety features, like not suspending a
 login shell, etc...
 
 
 >  For me, "suspend" is much too much to type, I prefer just "z"...
 
 That's another pet peeve of mine :), shell doesn't tab-complete
 functions (and/or aliases) in the initial position.  With completion
 it's "sus<TAB>", which is ok (and I don't want it to be too short
 anyway, lest I do it accidentally).
 
 But one can always do
 
     z() suspend
 
 :)
 
 
 > I don't think there is any good reason to build any of this
 > policy (name of the command, what signal is sent, ...) into
 > the shell when users can easily define their own.
 
 A lot of things can be left up to users, but I think it's fair to say
 that "suspend" is part of an established practice that a lot of people
 expect out of the box.
 
 
 -uwe