Re: master a154f0aa73b: shell-command-mode: New major mode for async-shell-command

Re: master a154f0aa73b: shell-command-mode: New major mode for async-shell-command

[Robert Pluim]

>>>>> On Sat,  1 Jun 2024 11:04:16 -0400 (EDT), Dmitry Gutov <dgutov@yandex.ru> said:

    Dmitry> branch: master
    Dmitry> commit a154f0aa73bceeaaeefcd243e7db9d09e3f450d7
    Dmitry> Author: Dmitry Gutov <dmitry@gutov.dev>
    Dmitry> Commit: Dmitry Gutov <dmitry@gutov.dev>

    Dmitry>     shell-command-mode: New major mode for async-shell-command
    
    Dmitry>     * etc/NEWS: Mention the additions.
    
    Dmitry>     * lisp/shell.el (shell-command-mode): New major mode (bug#71049).
    
    Dmitry>     * lisp/simple.el (async-shell-command-mode):
    Dmitry>     New variable, with default value pointing to that mode.
    Dmitry>     (shell-command): Refer to it here.
    Dmitry>     (async-shell-command): Update docstring.
    
    Dmitry>     * lisp/net/tramp.el (tramp-handle-shell-command):
    Dmitry>     Use the new variable when available.
    Dmitry> ---
 
    Dmitry> +(defvar async-shell-command-mode 'shell-command-mode
    Dmitry> +  "Major mode to use for the output of asynchronous `shell-command'.")
    Dmitry> +

Itʼs probably too late to change for emacs-30, but any particular
reason this is a `defvar' instead of a `defcustom'?

Robert
-- 

Re: master a154f0aa73b: shell-command-mode: New major mode for async-shell-command

[Dmitry Gutov]

On 12/09/2024 14:58, Robert Pluim wrote:
> Itʼs probably too late to change for emacs-30, but any particular
> reason this is a `defvar' instead of a `defcustom'?

A user option implies validity of different options, but as you can see 
the new major mode is there to fix a bug (a long-standing one, I guess), 
and the defvar is for any users who feel their workflows break because 
of that.

Not sure who would be the target audience, though.

Re: master a154f0aa73b: shell-command-mode: New major mode for async-shell-command

[Robert Pluim]

>>>>> On Fri, 13 Sep 2024 01:25:01 +0300, Dmitry Gutov <dmitry@gutov.dev> said:

    Dmitry> On 12/09/2024 14:58, Robert Pluim wrote:
    >> Itʼs probably too late to change for emacs-30, but any particular
    >> reason this is a `defvar' instead of a `defcustom'?

    Dmitry> A user option implies validity of different options, but as you can
    Dmitry> see the new major mode is there to fix a bug (a long-standing one, I
    Dmitry> guess), and the defvar is for any users who feel their workflows break
    Dmitry> because of that.

Well, the `defvar' also implies that some people might prefer the old
behaviour ☺️. But those people almost certainly know how to wield `setq'.

    Dmitry> Not sure who would be the target audience, though.

If you code it they will come.

Robert
--