In this chapter we’ve seen how to derive two complementary sets of navigational controls from a meta-tagged docbase. Tabbed indexes use metadata to shelve documents in flexible ways that can evolve as the data set grows. Sequential indexes enable users who’ve found an interesting shelf to scan backward and forward along its length.

We’ve explored these techniques in painstaking detail because, as I’ve said, the details matter enormously. There are really only two ways we find information in online systems—by navigation and by search. Many online systems are, to my way of thinking, sadly inadequate in both respects. Richly interconnected, multimode navigation, such as we’ve seen here, makes it a lot easier to find things. As a groupware developer, you should think of this as one way to reward users for the time they spend building docbases. People who know they’ll be able to find things are motivated to put things in. Whether the motivation is enlightened self-interest or the greater good doesn’t much matter; it’s the results that count.

Intelligent search results are another way to reward users who contribute to docbases. In the next chapter, we’ll see how to elaborate the results produced by any search engine into value-added views of a set of underlying docbases.