From xemacs-m  Fri Apr 25 13:07:43 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 NAA08605
	for <xemacs-beta@xemacs.org>; Fri, 25 Apr 1997 13:07:42 -0500 (CDT)
Received: from infodock.com (wave.infodock.com [206.13.40.192]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with ESMTP id LAA21297 for <xemacs-beta@xemacs.org>; Fri, 25 Apr 1997 11:07:00 -0700
Received: (from weiner@localhost) by infodock.com (8.7.4/8.7.3) id LAA02790; Fri, 25 Apr 1997 11:05:48 -0700
Date: Fri, 25 Apr 1997 11:05:48 -0700
From: Bob Weiner <weiner@infodock.com>
Message-Id: <199704251805.LAA02790@infodock.com>
To: xemacs-beta@xemacs.org
In-reply-to: <QQcmvd25186.199704251722@crystal.WonderWorks.COM> (message from Kyle Jones on Fri, 25 Apr 1997 13:22:11 -0400 (EDT))
Subject: Re: info mode mouse1

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

   KJ> Steven L Baur writes:
   >> Neal Becker <neal@ctd.comsat.com> writes:
   >> 
   >> > Have we changed info mode button1 to follow links, the same as
   >> > button2?
   >> 
   >> > It seems to be in 20.2b2.  I don't like it.
   >> 
   >> What's wrong with:
   >> 
   >> (add-hook 'Info-mode-hook
   >> (lambda () (remove-hook 'mouse-track-click-hook
   >> 'Info-maybe-follow-clicked-node)))

   KJ> I think you can leave it as is.  I checked and the mouse dragging
   KJ> functionality is there.  You can't set point inside a active
   KJ> region of text with the mouse, but I think that is minor.  There
   KJ> is the matter of consistency across various modes, but I don't
   KJ> know if that is worth tackling just now.

I was in full agreement with Kyle until this message, at which point,
I diverge a bit.  Mouse button 1 already has a wealth of consistent, global
functionality (setting point, selecting context-sensitive regions or
rectangles).  Adjusting it locally to select hyperlinks can only reduce this
consistency.  (I know the reasoning is likely for compatibility with web
browsers.)

I would vote for an option that a user would have to turn on to get this
behavior so that by default users who are familiar with other programs
besides web browsers, such as XEmacs itself, will not be bothered.

My personal experience in examining many local bindings of unmodified mouse
keys is that the bindings often do an imcomplete or inconsistent job,
sometimes even reducing functionality that exists in an available global
binding.

Finally, as anyone who has ever watched new users try a mouse interface
probably knows, they often click any button they can find, sometimes multiple
times until they get the result desired, and then they tend to repeat the
successful behavior because of the reinforcing feedback.  Thus, finding the
binding on the middle mouse button is a trivial exercise across the
population, whether or not they ever read a scrap of documentation.

Bob

