Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Katsumi Yamaoka]

>   * lisp/subr.el (y-or-n-p): Accept format string args.
>   * src/fns.c (Fyes_or_no_p): Accept format string args.

So, how do we write a Lisp source program that is for various
Emacs versions, and that uses `y-or-n-p' or `yes-or-no-p' like
the following?  (Please try evaluating it.)

(y-or-n-p "Is it 100% ok? ")

It's no more than an example; a prompt string may be the one that
a fickle program generates.

(y-or-n-p "Is it 100%% ok? ")

is not a solution because it's bad for Emacs 23 and earlier (and
XEmacs).  Taiji Can reported in the wl list that Wanderlust causes
an error recently as follows:

> Debugger entered--Lisp error: (error "Not enough arguments for format string")
[...]
>   y-or-n-p("Execute marks in %INBOX:username/clear@mailserver:143!!? (answer \"n\" to discard them) ")

Where "%INBOX:..." is a mail folder name that Wanderlust uses.
I think we can't help reverting it.

Regards,

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Chong Yidong]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

>>   * lisp/subr.el (y-or-n-p): Accept format string args.
>>   * src/fns.c (Fyes_or_no_p): Accept format string args.
>
> So, how do we write a Lisp source program that is for various
> Emacs versions, and that uses `y-or-n-p' or `yes-or-no-p' like
> the following?  (Please try evaluating it.)
>
> (y-or-n-p "Is it 100% ok? ")

Good point.  For compatibility, probably we should fall back on the old
behavior if only one argument is supplied.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Andreas Schwab]

Chong Yidong <cyd@stupidchicken.com> writes:

> Good point.  For compatibility, probably we should fall back on the old
> behavior if only one argument is supplied.

What's wrong with explicitly using format?

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Chong Yidong]

Andreas Schwab <schwab@linux-m68k.org> writes:

> What's wrong with explicitly using format?

Same reason as `message' and `error'.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Andreas Schwab]

Chong Yidong <cyd@stupidchicken.com> writes:

> Andreas Schwab <schwab@linux-m68k.org> writes:
>
>> What's wrong with explicitly using format?
>
> Same reason as `message' and `error'.

That does not answer my question.  Why could yes-or-no-p live without
format for more than 20 years?

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

RE: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-pand yes-or-no-p.

[Drew Adams]

> >> What's wrong with explicitly using format?
> > Same reason as `message' and `error'.
> That does not answer my question.  Why could yes-or-no-p live without
> format for more than 20 years?

1+  YAGNI.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Chong Yidong]

Andreas Schwab <schwab@linux-m68k.org> writes:

> Why could yes-or-no-p live without format for more than 20 years?

Presumably because no one bothered to change it.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Stefan Monnier]

> I think we can't help reverting it.

I tend to agree.  It doesn't seem important enough.


        Stefan

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Richard Stallman]

    Good point.  For compatibility, probably we should fall back on the old
    behavior if only one argument is supplied.

That will work in practice, but it is kludgy and ugly.  That is why I
didn't do this.  I regret that I did not initially made y-or-n-p and
yes-or-no-p use format.

Maybe the thing is to add two new functions which are similar
but always call format.
 
-- 
Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org, www.gnu.org

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Chong Yidong]

Richard Stallman <rms@gnu.org> writes:

>     Good point.  For compatibility, probably we should fall back on the old
>     behavior if only one argument is supplied.
>
> That will work in practice, but it is kludgy and ugly.  That is why I
> didn't do this.  I regret that I did not initially made y-or-n-p and
> yes-or-no-p use format.

Fair enough; I reverted the change.

> Maybe the thing is to add two new functions which are similar
> but always call format.

Probably not worth it, unless a much more concise set of function names
is offered.  I guess we'll have to live with this unfortunate historical
accident.

Re: /srv/bzr/emacs/trunk r102775: Allow format args for y-or-n-p and yes-or-no-p.

[Katsumi Yamaoka]

Chong Yidong <cyd@stupidchicken.com> wrote:
> Fair enough; I reverted the change.

Thank you!