bug#19467: 25.0.50; Default arguments for 'yes-or-no-p'

[Stefan Kangas]

tags 19467 + wontfix
close 19467
quit

Dani Moncayo <dmoncayo@gmail.com> writes:

> On Mon, Dec 29, 2014 at 10:32 PM, Drew Adams <drew.adams@oracle.com> wrote:
>>> I'd like I could choose the "yes"/"no" answer via M-n/M-p (as
>>> alternative to typing the answer directly in the minibuffer).
>>
>> FWIW, that defeats the purpose/intention of `yes-or-no-p'.
>> We have `y-or-n-p' for quick response.  The point of
>> `yes-or-no-p' is presumably to slow you down and make you
>> work a bit, so you take the question seriously. ;-)
>
> I agree wrt the intention of 'yes-or-no-p': force the user to think
> the answer well.  But IMO:
>
> 1. Looking for a default value with "M-n... RET" entails some work;
>    it's not something inmediate like a single keystroke (as in
>    'y-or-n-p').
>
> 2. The behavior I'm requesting feels quite natural to me: When Emacs
>    asks for input with a minibuffer and there are a few valid (or even
>    typical) values, those values are available in the minibuffer
>    "future history".

I think this would be a misfeature.  It would be too easy for a user
to answer the wrong thing in response to important questions such as
"Do you really want to delete this file?" by muscle memory M-p RET.

Additionally, as Drew points out above, we already have y-or-n-p which
is better for users that wish to answer these prompts quickly.  I for
one have used this in my init file for as long as I can remember:

(fset 'yes-or-no-p 'y-or-n-p)

I'm therefore closing this as wontfix.  If anyone disagrees with that,
please feel free to re-open.

Thanks,
Stefan Kangas