Research

Many of the user interfaces of current software force the user to do unnecessary work and waste his time, like these web pages. We have based this user interface solution on an unnecessary telephone metaphor, and you cannot even see, which buttons work and which ones don't; you are forced to try out all of them. The icons are obscure and all the browsers can't show the tooltips. And even if they could: why should we hide the structure of the information behind some arbitrary telephone-like buttons - even if it seemed neat?

In this interface, we have avoided long chains of pages - great. The buttons even show some feedback, but the feedback is inadequate. In general, using colors should always be a redundant coding method and, in this case, the color of the currently selected button is too hard to notice when you come to these pages first time. Our buttons work like tabs (good), but the visibility of the tab system is miserable. Users will find constructing an appropriate mental model of this system hard, because the user interface gives inadequate or misleading cues. In addition, unnecessary scrolling could be reduced by using vertically more compact navigation tabs on top of the page.

Good user interface solutions should be available in such a form that they could be implemented with minimal user interface design skills, quickly and cheaply. We have discovered that the same user interface problems, e.g. query formulation, visualization and management of hierarchies, and managing complex time intervals, tend to recur in various systems and contexts.

On these pages, we have used a feasible and usable abstract "tab" pattern in order to show five groups of information: GIST, people, research, teaching, and links, but we have ruined the good pattern by using some anti-patterns (an unsuitable metaphor, for example) and applying the good "tab" pattern incorrectly from an abstract representation into the concrete solution.

Bookmark Scroll Bar Demo - featured in CHI 2000 Interactive Posters