minibuffer helpful keybinding pointer obliterated

[Dan Jacobson]

Some commands tip you off that, e.g., upon typing M-x d i r e d RET
you get the helpful message that you indeed can run that command
with C-x d

1. And why doesn't this helpful message go into the *Messages* buffer?

Other commands don't produce the helpful message... but wait, they do,
but they get covered up real fast by other less helpful messages.

2. Consider typing M-x kill rectangle.
In the minibuffer we see

Loading rect...done

then the helpful message about the binding, which goes away after oh,
one second, replaced with

Loading rect...done

which then stays on the screen, much less helpfully than the
aforementioned helpful message, which, by its flighty nature, might
not be noticed therefore.

Yes, perhaps there is a variable that will keep it on the screen for a
whole 5 tortuous seconds or whatever, stopping all other actions
etc. but I'm telling you to get your goals straight as to what message
you want to _leave_ on the screen.

By the way, RMS sent a patch for my similar missive
<m21yhicey1.fsf@dan.jacobson.tw> but looking at the filename, I fear
it doesn't address today's issue.  I didn't test it because I fear
complexity.

RMS> Does this patch give you time to read the time-stamp message?
RMS> You'll need to evaluate the defun of basic-save-buffer
RMS> in Emacs (with C-M-x or M-x eval-region) before you try it.

*** files.el.~1.545.~   Fri Dec 21 17:52:10 2001
--- files.el    Sat Dec 29 21:36:59 2001
--- 2606,2612 ----....
-- 
http://www.geocities.com/jidanni/ Tel+886-4-25854780