Path: utzoo!attcan!uunet!lll-winken!lll-lcc!ames!amdahl!pacbell!att!mtunx!whuts!homxb!ho7cad!wjc
From: wjc@ho5cad (Bill Carpenter)
Newsgroups: comp.emacs
Subject: Re: Playing with the minibuffer -- two questions
Keywords: minibuffer
Message-ID: <338@ho7cad.ATT.COM>
Date: 15 Jun 88 12:47:20 GMT
References: <31024@yale-celray.yale.UUCP> <2379@mipos3.intel.com> <25724@bbn.COM>
Sender: nuucp@ho7cad.ATT.COM
Reply-To: wjc@ho5cad.att.com (Bill Carpenter)
Organization: AT&T Bell Laboratories
Lines: 16
In-reply-to: jr@bbn.com (John Robinson)

In article <25724@bbn.COM>, jr@bbn (John Robinson) writes:
>The minibuffer prompt and message strings (error output) are tucked
>away in C variables that aren't accesible to E-Lisp.  They could be
>with source changes.  But I don't see why this is eneded.

One reason it is  needed is  that sooner or  later, no matter how wide
your screen happens to be, you'll get an  important message that's too
wide to read.  By the time you can get  yourself  to the minibuffer to
make  it  multi-line,  the    message has been  squashed  (normally  a
convenience).  If you could get the message  into a normal buffer, you
could do whatever you wanted with it, including read it.>

(BTW, if somebody happens to know how to solve this particular problem
without  resorting to  "snatch-the-minibuffer", I'd sure  like to hear
about it.  Once looked for it in the manual, but without success.)

-- Bill Carpenter         att!ho5cad!wjc  or  attmail!bill