Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!pacific.mps.ohio-state.edu!zaphod.mps.ohio-state.edu!samsung!uunet!mstan!amull
From: amull@Morgan.COM (Andrew P. Mullhaupt)
Newsgroups: comp.misc
Subject: Re: Keyboard design (Was: Re: Xerox sues Apple!)
Summary: Not so fast there...
Message-ID: <626@s5.Morgan.COM>
Date: 24 Dec 89 19:35:33 GMT
References: <172@comcon.UUCP> <7326@ficc.uu.net> <9320@hoptoad.uucp> <9386@hoptoad.uucp>
Organization: Morgan Stanley & Co. NY, NY
Lines: 67

In article <9386@hoptoad.uucp>, tim@hoptoad.uucp (Tim Maroney) writes:
> In article <9375@hoptoad.uucp> tim@hoptoad.UUCP (Tim Maroney) writes:
> >>Keyboards *are* hard to learn.  Most people take a long time to learn
> >>to type with acceptable speed and accuracy.  But at least the keys that
> >>are of different type from the others (such as shift, control, etc.) do
> >>have distinctive shapes and locations, and all the keys have clear
> >>labels as to their function.
> 
> 
> Who ever said that keyboards were standardized?  You are bashing a
> straw man.  All you've done is *proven* my assertion -- that keyboards
> are hard to learn.  Even going from one to another is hard, to say
> nothing of learning them in the first place.
> 
Well learning a mouse is impossible from the 'touch typing' point of
view, because even if the idea of parking the rat in a 'home' position
on the pad occurs to you, the windowing software wants the position
of the mouse to indicate the active window. You can try picking up
the mouse and putting it in its 'garage' but this uses a  set of
muscles not intended by the ergonomic designers of the mouse; they're
not made to be easily lifted. Now if somebody used a mouse button for
'ignore motion', you'd get this effect, but this is only a poor
imitation. You'd have to actively return the mouse to 'home' after
most mouse-strokes, which significantly increases the annoyance 
of using it. 

Some GUI's pretend that mouse strokes and clicks provide a rich 
enough interface for most interaction. Unless you draw pictures all
day, I'd guess this is false. It is a simple truth of coding theory
that the keyboard with its far higher number of rapidly available
distinct inputs than a mouse is a broader channel for interaction,
perhaps excepting the case where pixel by pixel differentiation is
useful. (I.e. drawing). Anyone who has used the Logitech 'Point'
editor on the IBM PC can testify how trying to load up the mouse with
lots of signals --(three keys, chords, mouse direction strokes, screen
region commands, as well as 'point and shoot' box and button clicks.
Yikes! It even remaps the keyboard differently based on whether the
mouse is detected or not! If anyone wants to know how NOT to build a
user interface - get this product!)-- leads to an overwhelming desire
to not use the rat. Mercifully, some environments like 'SunView' are
really just a bunch of keyboard oriented windows on screen at the
same time; so the rat is not essential in all GUI's. 

On the other hand, if what you need is input of planar x,y coordinates
with medium resolution, then the mouse can be useful. That same
Logitech mouse interface product (with only up - down, right - left,
and each of three buttons mapped to single keystrokes,) adds some value 
to Lotus Manuscript's previewing screen, where a variable magnification
box is dragged around a page of previewed text. This accounts for about
5% of your time using the product, though, and Lotus probably made the
right choice leaving the mouse out of their product since there was no
single standard mouse interface they could expect on the IBM PC.

I think for reasons of speed and comfort, the mouse is a poor substitute
for a command line with a nice command recall/editing feature. It is
not easy to use it without looking for it after you've gone to the
keyboard and vice versa. Trackballs fixed to the keyboards can alleviate
some of this problem, but the actions needed to push a mouse or roll
a trackball are tiring compared to using a good keyboard, mostly
due to the intermittent moving the hand to and from the keyboard. If
you're like me and most of your input to the computer is text, then a
pointing device, and an environment tailored to one is a waste of time.
The only parts I read in the GUI manuals these days are the parts
relevant to the possible of escape from them.

Later,
Andrew Mullhaupt