Path: utzoo!utgpu!watmath!clyde!att!rutgers!mailrus!bbn!mit-eddie!uw-beaver!rice!sun-spots-request
From: steinmetz!grymoire!barnett@uunet.uu.net (Bruce Barnett)
Newsgroups: comp.sys.sun
Subject: Re: Warping mouse cursors
Message-ID: <8901032221.AA00649@grymoire.steinmetz.GE.COM>
Date: 11 Jan 89 07:06:22 GMT
Sender: usenet@rice.edu
Organization: Eikonix Corp., Bedford, MA
Lines: 73
Approved: Sun-Spots@rice.edu
Original-Date: Tue, 3 Jan 89 17:21:25 EST
X-Sun-Spots-Digest: Volume 7, Issue 94, message 6 of 13

Ye Moderator says.
>On my 3.2 system.....

Yup. Sorry for the mistake. I've been working too hard. Is my boss reading
this?

Bill also says:

>So I quit this textedit window and confirm the quit before the alert
>box shows up.  Then I start moving the mouse toward the mailtool icon to
>read some mail that just came in.  While I am moving the mouse, will it
>jump halfway across the screen, moves a bit, then jump back?  Or is it
>smart enough to say "already confirmed---no warping needed"?

If you click the left mouse button before you move the mouse, which
admittedly must be learned, it warps to the alert box, and warps back
right away. If you move the mouse after performing an operation that
generates an alert, you will be moving it relative to the warped mouse
position, which is where the trouble lies - it ends up in an unexpected
position.  This is aggravating if you didn't expect an alert. Perhaps the
Alert package should warp the mouse to the dialog box, and prevent the
mouse from leaving the box until the alert is dismissed.  At this point,
the mouse is warped back to the original location.

As an aside, someone said the warping mouse cursor can be controlled by a user
preference option.

[[ I would also like to be able to ignore the alert if I want to.  I don't
know if I can currently, but your scheme of limiting the cursor motion
would certainly preclude that.  --wnl ]]

Re: warping with the Open Look scrollbar:
>I would contend that the elevator is poorly designed if it needs
>warping to make it work correctly.  What's wrong with a stationary button
>that scrolls forward and another that scrolls backward (buttons that are
>easier to find than SunView's, that is)?

SunView's current scrollbars are not intuitive. Remember that a consistent
user interface is important to new users. The function of the left mouse
button is to poke at devices/buttons. (This image comes to mind of a
primative human poking at some alien device with his/her forefinger :-)

I would say the most common actions a scrollbar should provide is:

	go to beginning 
	go to end
	go forward one line (can be repeated)
	go back one line (can be repeated)
	go forward one screen (can be repeated)
	go back one screen (can be repeated)
	go to a particular spot.

If you look at sunview, the only "intuitive" action is scrolling down one
line, which can be repeated. The rest of the operations must be learned by
trial and error, or by a manual or tutorial.

Now if you look at Open Look, you will see that ALL of these operations
can be performed with the left mouse button.  Only two functions are in
static locations. the rest of the "buttons" move with the "elevator".

Now if you had to provide ALL of these options from the left mouse button
without warping the cursor, you would have to make room for three buttons
at each end. Then how would you know intuitively which function each
button had?

Perhaps someone can come up with a better design.....

Also, since the mouse is only warped within the scrollbar area, it isn't
TOO disturbing. IMHO.  [[ Admittedly, it is the least offensive example of
mouse warping that I have seen.  In fact, it's kind of clever.  --wnl ]]

Bruce G. Barnett 	barnett@ge-crd.ARPA, barnett@steinmetz.ge.com
			uunet!steinmetz!barnett