The Macintosh GUI is a shadow of what it could be

Written by David Tebbutt, MacUser Jan 1989 (guess) - scanned

The Macintosh user interface is great, isn't it? It's the best on the market, yet it actually offers too much for some people and too little for others. 'How can it possibly offer too much?' I hear you ask. The answer lies in the fact that it exploits three of our senses: sight, touch and sound. What if you lack one of these? In particular, what if you're blind? Try getting the Mac to run your favourite program while you've got your eyes shut, you'll soon understand what I mean.

This isn't the sort of issue that would normally cross my mind, but I recently went to help a blind man sort out some computer problems. It was there that I learned what a frustrating experience computing can be for someone who can't see what's going on. On one hand, the computer allows this man to write books - something he would have found very difficult otherwise. On the other hand, he's using a wordprocessor specially written for blind people and this, quite unforgivably, is inconsistent and contains bugs, some of which cause him to lose his current document files.

The trouble is that the authors of the wordprocessor don't earn enough from their limited marketplace to enable them to fix the problems. The bugs and anomalies which cause the blind man's problems aren't going to go away. He needs a text-only wordprocessor with simple controls so that his speech synthesiser can tell him what's on the screen. This device has no problems with characters and words, but it falls apart when it encounters a typographical or control character. Imagine what it would make of the Macintosh screen.

The man is now the owner of an industry standard, although old-fashioned, wordprocessor which uses the same controls as the buggy one and which has no peculiar graphics to fox the synthesiser. I'm hoping that, by going the mass-market route, he now has a well proven product which won't eat his files or behave inconsistently.

Although I'm sure that people have written special Macintosh programs for the blind, I would say that for such a person, the Macintosh user interface is too good.

At the other end of the spectrum, just over a year ago, I met Bill Buxton, a scientist at the UK operation of Xerox Palo Alto Research Center, EuroPARC in Cambridge. The American end of PARC is where the Macintosh style of user interface was first invented. While the blind man had made me realise that there are advantages to a strictly character-based approach to computing, Buxton opened up new user-interface possibilities for people whose senses of hearing, sight and touch remain intact. He believes the Macintosh user interface should be treated as only the beginning, rather than the end point of user-interface design.

He was quite amused that all the computer manufacturers are trying to emulate a user-interface style which was first developed in 1974. He thinks that by the 1990s we will be regarding the Macintosh in the same light that we currently regard the COBOL programming language. Since the '90s begin next year, he's probably wrong about the timescale, but it's an interesting thought. Of course, a research institute like PARC wouldn't be seen dead using COBOL. It writes in its own language, Smalltalk, or in a Smalltalk inspired adaptation of LISP. I suppose COBOL could be considered old-hat.

So how is the Mac going to become 'old-hat' in such a short time? The answer, according to Buxton, is that users will demand improvements to the user interface if they know what's possible. Manufacturers today, and Apple is by no means the worst example, hide behind a smoke screen of 'But there's no user demand.' At the time I saw Buxton, MacPaint still insisted you use the 'hand' to move your paper. This meant stopping in mid-drawing, selecting the hand, moving the paper, getting back your painting tool and picking up where you left off.

Buxton figured that a far more natural solution was to activate a touch tablet with your left hand moving the paper directly to the desired position. This way, your right hand remains in control of the painting tool. Details of such a device were published in 1980.

He also suggested that far more use could be made of gestural control of a computer. He showed an example of ringing some items on the screen and running the line to another location. As soon as he let go of the mouse button, the circled items moved to the new location. In another example he trailed the line across the marked area, this deleted the items encircled. These gestures would come naturally to a writer or editor, and could well reduce the number of keystrokes involved in using a wordprocessor. This idea was first outlined in 1957 and a working version written in 1969.

Imagine what you could do with a mouse that responded to variable pressure. You could change the speed at which you scrolled around a document or spreadsheet, or you could vary brush width simply by increasing the downward pressure on your mouse. If you've got colour, you could change the intensity of colour with mouse pressure. When working with sound, you could vary the loudness.

Even these few examples show that there's much scope for improvement in the way we interact with our computers. I hope it's enough to get you thinking about how your experience with the Macintosh can be improved. Make sure that your proposals improve both useability and usefulness of the machine and tell Apple what you think.

We must never give Apple justification for the excuse 'But there's no user demand'.