Bad UI

Future tech is a major part of many science fiction stories. Indeed, if the whole point of the story is to examine the consequences of a novel technology, then the technology is going to be highly visible.

Something I like to think about is what the user interface would be. This is the first in an occasional series on user interfaces.

Every tool has a user interface

A user interface is how the user tells the tool what to do. Every tool we use has a user interface, or UI, of some kind: the UI for a hammer is to hold the handle with the heavy end pointing up and then to strike the heavy end against the object needing percussive force.

Physical tools tend to have tactile and relatively obvious usage modes: hammer, saw, and screwdriver all have pretty simple interfaces although there is a skill and subtlety involved in using such tools well. And these tools often have alternative usage modes: remove nails with a claw hammer; open a paint tin with a flat head screwdriver; or play a saw as a musical instrument.

Tools which use mechanisms have more complex and less obvious interfaces. The trigger switch of the gun or the drill is pretty clear and simple but it could have been done another way. The handlebars of an upright bicycle serve the dual purposes of steering and stability and so would be difficult to arrange much differently, whereas the arrangement of controls in a car is largely by convention. But again, there is subtlety to be learned in controlling these mechanisms well.

All of these systems have a direct relationship between the controls and the effect, though. This relationship breaks down for electronic devices, and especially digital electronics: software has no physics. The basic problem is that because (to a first approximation) computers can do everything it is hard to figure out how to tell them to do anything.

Too many features

The Guardian published this article recently, bemoaning the meaningless proliferation of features in household appliances.

The premise of the article is that there are too many useless features on these appliances, but I contend that the real problem is inadequate UI – the user has to know that all of those features are available and then make a conscious choice to use each one. A better UI would provide options appropriate to the situation, or based on previous usage patterns. You’ve put a piece of bread in your toaster? The SmartToast would present a big button to say “toast this how I like it”, with smaller buttons to choose whether to burn in a design of your choice: weather forecast, smiley face, Darth Vader helmet. It’s a waffle instead? SmartToast sets the crispness level to your usual choice.

And so on. If you don’t want that level of complexity, you could choose to just have the toaster present something closer to a simple “heat for this long” control, but with the added consistency of taking account of whether the heating element is already hot or not.

The point here is that the options presented to you should be appropriate for the circumstances, rather than necessarily everything at once. This is the thing that Apple’s UI people understand (mostly, iTunes notwithstanding) and Microsoft’s historically do not (the Office ribbon was theoretically an improvement but broke every existing user’s established usage patterns).

Next time I want to talk about feature hiding.

Leave a Reply