Re: lisp-complete-symbol is too loquacious in minibuffer

[Stefan Monnier]

>>> Stefan, there's a bug in your latest changes to this function: when a
>>> completion has been found, the *Completions* buffer is buried but its
>>> window remains visible.  In previous versions, the window was deleted
>>> as well.
>> 
>> If you do it the other way around you also have a bug in that it may
>> remove a window that was there before.

> It seems the only good way to handle all cases is preserving the
> window configuration before first M-TAB, and restoring it afterwards.

The *Completions* buffer may be displayed in a different frame.


        Stefan