Abstract
Thankfully, the discipline of software development has evolved away from the original "waterfall "model - design, implement, deploy - and moved to the prototyping model - design, use, re-design, re-useS, deploy. But Paul Pangaro thinks that's only a half-way point. He says that HCI must recognize that the end-user is also constantly prototyping - that is, iteratively converging upon specific goals, not just executing pre-defined tasks. Why doesn't the software help?
Sure, search engines let us refine queries, but they don't help us evolve an understanding of what we find - which presumably is why we search. Word processors allow easy editing of text, but don't provide tools for adding rigor to that text - conciseness, completeness, and consistency, which are the goals of the editing process.
In developing user interfaces for animation, content management,
browsers, and search tools, Pangaro has focused on features that
encourage end-users to be prototypers in the context of their
own process. With examples from work at pre-Media Lab MIT, during
the heights of AI mania, and now in HTTP and HTML, he proposes
ways of minimizing the clumsy interactions we suffer every day
at the screens of our browsers, using dumb search engines and
someone else's taxonomies.