Go to the first, previous, next, last section, table of contents.

Alan Kay

In the late 60s, Alan Kay was a wild-eyed maniac doing his thesis on The Reactive Engine and forecasting that in the near future we would have cheap personal laptop computers with crisp color graphics and more computing power than an IBM mainframe. Obviously a nut case.

Today we in fact have cheap personal laptop computers with merely stunning color graphics, and so much more computing power than an IBM mainframe of that era as to make the comparison ludicrous: In retrospect, Alan was off mainly in being too conservative.

We owe much of the utility of those laptops to Alan Kay: When forecasting the arrival of what he called the Dynabook, he also observed that they would scarcely be useful if equipped with software no better than the IBM mainframe operating systems of the time, and he set out to remedy the problem:

He convinced the Xerox Palo Alto Research Center to fund a lab to work on the problem, put together the best hardware prototype of the Dynabook that he could manage with available technology -- the Alto, which with bitmapped display, mouse and network connectivity was in many ways the prototype for the modern workstation -- and set out to design software to make it all usable.

He chose schoolchildren as his test audience early on, observing that children need much better hardware and software than adults, who can successfully be paid to put up with almost any degree of awfulness, and did a series of cycles of design and implementation followed by analysis of people actually using the system and back to the drawing board.

The result was Smalltalk, which did much to mainstream object-oriented programming, and a snazzy mice-and-menus user interface which Xerox, in its inimitable fashion, commercialized in the form of the Xerox Star computer, a runaway commercial failure. (Xerox is very good at not making money selling computers, having a long series of such efforts, stretching back to the Sigma V computer. Hmm? Never heard of it?) All was not lost: A little company named Apple was looking for something exciting to do for their next computer, toured Xerox PARC, and the rest is history. (Well, actually, their first attempt was the Apple Lisa, which in commercial terms was right on a par with the Star, but the second try was the MacIntosh -- first microcomputer computer to ship with no end-user programmability whatever -- and things turned out ok in the end, so much so that Apple eventually sued Microsoft for stealing Xerox's ideas from them.)

My favorite Alan Kay quote:

Simple things should be simple. Complex things should be possible.

I consider this a deceptively subtle insight on user interface design: Too many interfaces try to make complex things simple -- which is combinatorially impossible -- and succeed in the end only in making simple things complex.

I'm also fond of Alan Kay's observation (during a talk at the UW) that "Graduate students are like geese: We imprint on the first good idea we see, and spend the rest of our careers chasing it."

And his observation that the great problem with Lisp is that it is just good enough to keep us from developing something really good. (When everyone laughed, he added plaintively, "It's true! You know it's true...)

Go to the first, previous, next, last section, table of contents.