Afterword: Dealing with the Inmates

Since About Face was first published in 1995, the world of digital products has certainly changed. The Web changed the way we think about information access, and the dotcoms have come and gone. As per Moore's Law, hardware has gotten an order of magnitude faster, and storage two orders of magnitude larger. Handheld data devices, then a cumbersome curiosity, are now everywhere and are beginning to merge with wireless telephones and digital cameras.

But for all the steaming piles of glorious technology around us, our lives are not significantly better. Computers are still hard to use. Cameras and phones sometimes crash and reboot instead of making calls and taking pictures when we'd like them to. And now our DVD players blink 12:00 after power loss along with our VCRs.

When will we learn that it isn't about the technology, but rather about the behavior of these products? Why should anything be different just because of hardware and bandwidth improvements? The old problem of stupid, rude, inappropriate software and software-enabled products still exists, same as ever.

Ultimately, we will make digital products better by examining and satisfying the user's goals. We will not make them better by moving to new platforms or by improving technology. Our technology is superb. What it lacked in 1995, and still lacks today, is better consideration of the human.

What is required is more than just great design. What we need are product development organizations that foster great design. Designers need to become not only advocates for the user, but also advocates for change within their organizations. Old thinking, based on old assumptions about technology constraints, permeates many development organizations. It is something that practitioners must actively take steps to overcome.

Scarcity Thinking

We now have a quarter-century of refinements to the PARC paradigm. We know how to create good error messages, confirmation dialog boxes, and alerts. But we still don't have much experience creating rich, visual, unified interfaces that work hard to support users. We have years of experience building systems with robust data integrity and sophisticated hierarchical file systems, but we don't yet have experience creating systems with data immunity and attribute-based storage systems.

The problem is simple: Everything we know about computers is wrong! Fifty years ago, there was less computing power on the entire planet than is in your wristwatch today. There is more computer power in your family car than there is in the space shuttle. Just twenty-five years ago, computers were precious commodities: extremely expensive, limited, and weak. In 1975, a state-of-the-art IBM 370/135 mainframe had 144K of main memory. It had two 100 MB hard-disk drives, each the size of a refrigerator. It had a card reader, a punch, and a chain printer. It resided in its own room, nestled deep inside its own building.

Computing resources then were always very scarce. There was never enough memory, never enough storage, never enough cycles, and never enough bandwidth. If you wanted to be good at what you did, you worked hard to maximize the scarcest resource — you made sure that the CPU got all the breaks. You developed systems to maximize the use of disk, of RAM, even of punched cards.

The senior programmers and computer scientists in business and academia who lived through that era learned to think in terms of this rarified access to computing resources: scarcity thinking. And these same men and women, who know deep down in their guts that computer resources are scarce, are still setting the pace in the software industry today.