From xemacs-m  Wed Feb 26 20:45:09 1997
Received: from venus.Sun.COM (venus.Sun.COM [192.9.25.5])
	by xemacs.org (8.8.5/8.8.5) with SMTP id UAA19366
	for <xemacs-beta@xemacs.org>; Wed, 26 Feb 1997 20:45:08 -0600 (CST)
Received: from Eng.Sun.COM ([129.146.1.25]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id SAA17605; Wed, 26 Feb 1997 18:44:38 -0800
Received: from kindra.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3)
	id SAA19768; Wed, 26 Feb 1997 18:44:31 -0800
Received: from xemacs.eng.sun.com by kindra.eng.sun.com (SMI-8.6/SMI-SVR4)
	id SAA09581; Wed, 26 Feb 1997 18:44:31 -0800
Received: by xemacs.eng.sun.com (SMI-8.6/SMI-SVR4)
	id SAA27228; Wed, 26 Feb 1997 18:44:31 -0800
Date: Wed, 26 Feb 1997 18:44:31 -0800
Message-Id: <199702270244.SAA27228@xemacs.eng.sun.com>
From: Martin Buchholz <mrb@Eng.Sun.COM>
To: Kyle Jones <kyle_jones@wonderworks.com>
Cc: xemacs-beta@xemacs.org
Subject: Re: balloon-help 1.04
In-Reply-To: <QQcemx15741.199702260349@crystal.WonderWorks.COM>
References: <QQcemn13413.199702260126@crystal.WonderWorks.COM>
	<m3iv3gl0xi.fsf@jens.metrix.de>
	<QQcemx15741.199702260349@crystal.WonderWorks.COM>
Reply-To: Martin Buchholz <mrb@Eng.Sun.COM>
Mime-Version: 1.0 (generated by tm-edit 7.100)
Content-Type: text/plain; charset=US-ASCII

>>>>> "Kyle" == Kyle Jones <kyle_jones@wonderworks.com> writes:

Kyle> Proportional width fonts still need work.  The frame is not sized
Kyle> properly for them, because there's no easy way to do it.  In fact
Kyle> I don't know if there's any way to do it except by binary
Kyle> searching to the correct size, which is slow.  If I had a
Kyle> current-pixel-column function, I could put the cursor on the end
Kyle> of the longest line, call current-pixel-column and then size the
Kyle> frame to fit.  I want this function.  I ardently desire this
Kyle> function.  Would someone please write it?

I agree.  However, you'll have to actually check every line, not just
the char-longest one, since the pixel-longest might not be the
char-longest.  The CJK language users have it bad, since their
characters frequently exhibit 2-1 ratios even in a fixed-width font.
For them we currently kludge things by using (charset-columns) to
approximate the display-width.  This works well as long as they use
fonts that have been chosen to go together well.

For a general solution, we will pay a performance penalty.
(previous-line), for example should go (maybe?) to the spot as close
to its starting pixel position as possible.  But this would result in
(previous-line) possibly moving to a different character depending on
the current frame and device.  We don't want that.  

It's a difficult problem.

Martin

