Bug with icomplete-mode

Bug with icomplete-mode

[Slawomir Nowaczyk]

Hello,

I have recently switched from (now obsolete) iswitchb to icomplete-mode
and I have encountered an issue with that.

emacs -Q
(icomplete-mode t)
(setq icomplete-show-matches-on-no-input t)
(get-buffer-create "aaa")
(get-buffer-create "bbb")

Press C-x b (switch buffer)
C-. (select buffer "aaa")
C-j

Emacs switches to "*Messages*", not the "aaa" buffer I have selected.

As far as I can tell, it's because of this part in the minibuffer-force-complete-and-exit
  (if (and (eq (minibuffer-prompt-end) (point-max))
           minibuffer-default)
      ;; Use the provided default if there's one (bug#17545).

-- 
 Best wishes,
   Slawomir Nowaczyk
     ( Slawomir.Nowaczyk@hh.se )

Program:  A magic spell cast over a computer allowing it to turn
one's input into error messages.

Re: Bug with icomplete-mode

[Dmitry Gutov]

On 05/16/2015 02:49 PM, Slawomir Nowaczyk wrote:

> Emacs switches to "*Messages*", not the "aaa" buffer I have selected.

Yep, looks like a problem. But next time please do M-x report-emacs-bug.

> As far as I can tell, it's because of this part in the minibuffer-force-complete-and-exit
>    (if (and (eq (minibuffer-prompt-end) (point-max))
>             minibuffer-default)
>        ;; Use the provided default if there's one (bug#17545).

And this part is due to the bug that is referenced in the comment.

Looking at it, it seems that a better way to resolve it would be to 
always make the default value the first item, even when it's not in the 
collection passed to completing-read.

Re: Bug with icomplete-mode

[Stephen Leake]

Dmitry Gutov <dgutov@yandex.ru> writes:

> On 05/16/2015 02:49 PM, Slawomir Nowaczyk wrote:
>
>> Emacs switches to "*Messages*", not the "aaa" buffer I have selected.
>
> Yep, looks like a problem. But next time please do M-x report-emacs-bug.
>
>> As far as I can tell, it's because of this part in the minibuffer-force-complete-and-exit
>>    (if (and (eq (minibuffer-prompt-end) (point-max))
>>             minibuffer-default)
>>        ;; Use the provided default if there's one (bug#17545).
>
> And this part is due to the bug that is referenced in the comment.
>
> Looking at it, it seems that a better way to resolve it would be to
> always make the default value the first item, even when it's not in
> the collection passed to completing-read.

+1
-- 
-- Stephe