From xemacs-m  Mon Dec 23 16:59:47 1996
Received: from alphatech.com (erebus.alphatech.com [198.112.236.2])
          by xemacs.cs.uiuc.edu (8.8.4/8.8.4) with SMTP
	  id QAA22674 for <xemacs-beta@xemacs.org>; Mon, 23 Dec 1996 16:59:42 -0600 (CST)
Received: from venus.alphatech.com by alphatech.com (4.1/SMI-4.1)
	id AA26246; Mon, 23 Dec 96 17:55:46 EST
Received: from euler.alphatech.com by venus.alphatech.com (4.1/SMI-4.1)
	id AA08248; Mon, 23 Dec 96 18:02:29 EST
Received: by euler.alphatech.com (5.x/SMI-SVR4)
	id AA03702; Mon, 23 Dec 1996 17:54:47 -0500
Date: Mon, 23 Dec 1996 17:54:47 -0500
From: greg@euler.alphatech.com (Greg Klanderman)
Message-Id: <9612232254.AA03702@euler.alphatech.com>
To: David Moore <dmoore@UCSD.EDU>
Cc: xemacs-beta@xemacs.org
In-Reply-To: David Moore's message of 23 Dec 1996 11:57:51 -0800
Subject: Re: End of Buffer Tags
Reply-To: greg@alphatech.com
References: <9612231735.AA00669@euler.alphatech.com>
	<m2pw01m93d.fsf@altair.xemacs.org> <rvafr59gow.fsf@sdnp5.ucsd.edu>



>>>>> "David" == David Moore <dmoore@UCSD.EDU> writes:
David> 
David> Steven L Baur <steve@miranova.com> writes:
Greg> While we're on the subject, another thing I've always found
Greg> annoying occurs when you're entering something in the minibuffer
Greg> and you exceed the width of the screen.  It scrolls left by a
Greg> full screen and all you see is the cursor, with no context.  If
Greg> it only scrolled like 1/4 screen at a time it would be much
Greg> nicer.  If there's already a package or variable to do this let
Greg> me know - maybe it should be the default?
>> 
>> There's rsz-minibuffer.  There's code in the sample.emacs to turn it
>> on.  There are problems with some of the settings (see recent postings
>> on comp.emacs.xemacs), but the settings:
>> 
>> (cond (running-xemacs
>> (autoload 'resize-minibuffer-mode "rsz-minibuf" nil t)
>> (resize-minibuffer-mode)
>> (setq resize-minibuffer-window-exactly nil)))
>> 
>> mostly does a good job.

I checked this out and its definitely a solution, though not really
the one I was looking for.  Instead of having the minibuffer grow to
multiple lines I was hoping for something that would scroll the one
minibuffer line to keep as much as possible visible when it is wider
than the screen, like most single-line-of-text entry widgets.

I'll use it a bit and see if it grows on me.

David> 
David> 	I'll have to look at this.  However, the problem I've seen needs
David> to be fixed in the ''default'' configuration.  For example, when
David> movemail reports an error under Gnus, I often get stuck a yes/no prompt
David> where it seems to be _impossible_ to see the message I am being prompted
David> for.  The minibuffer is scrolled all the way to the point where I insert
David> text.  And any cursor movement command doesn't show me the line, since
David> I'm already at the beginning of the insertion area.  I've tried clicking
David> things, using ^L, trying moving.
David> 
David> 	Am I missing something, or can I just not see these messages?
David> This is with default single attached minibuffer.  At least in the text
David> window it grows the minibuffer so you can see messages.

Have you tried `scroll-right' and `scroll-left' (C-x > and C-x <)?
These work for me.  I have keys bound to scroll left and right by 1/4
screen at a time which works even better but is still not an ideal
solution.  Having to know to press "C-x >" to find out what emacs is
telling you is not really something a novice is likely to figure out.

Greg

