From xemacs-m  Fri Mar 21 15:39:27 1997
Received: from alphatech.com (erebus.alphatech.com [198.112.236.2])
	by xemacs.org (8.8.5/8.8.5) with SMTP id PAA18682
	for <xemacs-beta@xemacs.org>; Fri, 21 Mar 1997 15:39:26 -0600 (CST)
Received: from venus.alphatech.com by alphatech.com (4.1/SMI-4.1)
	id AA12322; Fri, 21 Mar 97 16:34:06 EST
Received: from pochacco.alphatech.com by venus.alphatech.com (4.1/SMI-4.1)
	id AA25279; Fri, 21 Mar 97 16:42:19 EST
Received: by pochacco.alphatech.com (SMI-8.6/SMI-SVR4)
	id QAA01506; Fri, 21 Mar 1997 16:31:47 -0500
Sender: greg@pochacco.alphatech.com
From: greg@alphatech.com (Greg Klanderman)
To: Raymond Toy <toy@rtp.ericsson.se>
Cc: XEmacs Mailing List <xemacs-beta@xemacs.org>
Subject: Re: Scrollbars and large inline images
References: <1266.857430142@rtp.ericsson.se>
Reply-To: greg@alphatech.com
Mime-Version: 1.0 (generated by tm-edit 7.103)
Content-Type: text/plain; charset=US-ASCII
Date: 21 Mar 1997 16:31:45 -0500
In-Reply-To: Raymond Toy's message of Mon, 03 Mar 1997 18:02:22 -0500
Message-Id: <ugg1xp3pa6.fsf@pochacco.alphatech.com>
Lines: 29
X-Mailer: Gnus v5.4.11/XEmacs 19.15


>>>>> several weeks ago "Raymond" == Raymond Toy <toy@rtp.ericsson.se> wrote:
Raymond> While reading Georg Nikodym's screen shot, I noticed that the Lucid
Raymond> scrollbars work quite well.  I could scroll the up the buffer one line
Raymond> at a time and the inlined image would have an extra lines worth of
Raymond> image to show.  Of course, when there were no more text lines to be
Raymond> shown, pressing the down arrow on the scrollbar jumped to the "next"
Raymond> line and showed the bottom of the image.
Raymond> 
Raymond> So, for the most part the scrollbar did the right thing.  What I'm
Raymond> wondering is if it were possible for the scrollbar to scroll by a
Raymond> nominal amount even if the apparent line height was very large.
Raymond> Something like if nominal-line-height set, then scroll image by the
Raymond> nominal-line-height.  Otherwise, scroll by physical lines.  The
Raymond> default being nominal-line-height being unset.

I think this would be a wonderful thing, the inability to see all of
an image and have it scroll sanely is a big problem.  However, after a
cursory look through window.[ch] I imagine this would require major
changes.  Looks like the window's notion of where in the buffer to
start displaying is just a character position (no surprise there), so
you must display starting at the top of a line, no matter how high the
line is.

I wonder if tm could display inline images by breaking them up into
many nominal-height lines?  If the interline space can be set to zero,
that's likely to mostly solve the problem.

Greg

