On 23 December 2016 at 18:59, martin rudalics <rudalics@gmx.at> wrote:

> > ​I'm not trying to override the user here, I'm just trying to avoid
> adding
> > another configuration option.​
>
> You change long-standing behavior when you add such an option.


​I haven't talked about adding an option.​


> Did
> ​ p​
> eople agree that the buffer should be preferably not displayed?  The
> current status is that
>
>   To run COMMAND without displaying the output
>   in a window you can configure `display-buffer-alist' to use the action
>   `display-buffer-no-window' for the buffer `*Async Shell Command*'.
>
> With your proposal the buffer would not be displayed and the user would
> have to delete the ‘display-buffer-alist’ entry to display the buffer.
>

​That's the case with my original patch. But I agreed with Eli (and you)
that the default behaviour should be the same as at present.​

​I'm just asking how to implement this, preferably without adding another
configuration option.​


> > It would be nice if the user only had to change one thing to enable
> hiding
> > the async output buffer until there is output.
>
> But when you add the entry to ‘display-buffer-alist’ the buffer will
> already be hidden without any user intervention.


​Again, you are describing the behaviour of my original patch. ​I then
suggested adding an option to display-buffer-alist's defcustom
specification. But you said that is not allowed (if I understood correctly).

I will try to outline the current position again:

1. I agree that the current default behaviour should remain as it is.

2. I would like to add the ability to easily turn off displaying the Async
Command output buffer until there is some output.

3. I suggested achieving this by i) adding an option to
display-buffer-alist, and ii) documenting this in shell-command.

4. You stated that it is not allowed for code to change
display-buffer-alist. I was puzzled by this, because other user variables
such as auto-mode-alist are changed by code as well as by the user. In any
case, I am not suggesting automatically changing display-buffer-alist;
rather I have suggested adding an option to the customization menu for
display-buffer-alist.

So, maybe you could give your opinion of my current suggestion (3),
ignoring my original patch?

Sorry if I have confused you, and I hope the above makes things clearer.

-- 
http://rrt.sc3d.org