Logo IMG


Qwerks of History

Brian Hayes

Qwerks to Come

The premise of this column is a bit of fraud. It's too easy to look back 35 years and pick out a few ideas that proved to be winners; the real trick is to name the present-day technologies that will emerge as qwerks of the future. I'm not even sure about the further evolution of the three choices discussed here. In one form or another the Internet will surely survive, and it's a fair bet there will be microprocessors that recognize x86 instructions for some time to come. But what will systems software look like in 10 years, or 35?

A number of visionary proposals are on the record already. Jef Raskin, the first architect of the Macintosh, offers the idea of a "zooming space" to replace operating systems, Web browsers and various other kinds of programs. The metaphor is geographic: Documents are spread out over a landscape, and the user soars above them, zooming in to see details and zooming out for an overview. It's an appealing (if slightly dizzying) notion, but I wonder whether it will scale any better than nested trees of directories.

David Gelernter of Yale argues that the key organizing principle should be temporal rather than spatial. His "lifestreams" model arranges documents in sequences and subsequences, with various facilities for browsing and searching. Underneath this interface is a database called tuple space. It's worth noting that both Raskin and Gelernter focus on the problem of organizing documents—texts, music, pictures, or more generally what nowadays tends to be called "content." But the original purpose of operating systems was mainly to organize the software that runs the computer itself, and this still needs to be done. Can a zooming space or a lifestream help me manage that gargantuan TeX tree?

The most radical proposal comes from Donald A. Norman, another interface expert who was once at Apple. He solves the complexity problem at a stroke: If computers are too hard to use, just get rid of them! They can be replaced by more-specialized "information appliances"—one device to send e-mail, one to balance your checkbook, one to do tax returns, etc. I don't trust myself to gauge the plausibility of this idea because I find it so thoroughly uncongenial. I want to know which information appliance replaces the computer that I compute with.

As for my own predictions, I'm going to paraphrase an old joke that used to be told about the Fortran programming language. I don't know what the operating system of the future will look like. All I know is that it will be called Unix.

© Brian Hayes

comments powered by Disqus


Of Possible Interest

Computing Science: Belles lettres Meets Big Data

Technologue: Quantum Randomness

Technologue: The Quest for Randomness

Subscribe to American Scientist