Bryan and I were commenting yesterday how we use the Index feature in Visual Studio very little…well actually we use it indirectly a lot. When we’re programming we hover for details on functions or press F1 to get more information on routines that the cursor is over. However, if we’re going to search for something by typing, we just go to Google.
I know many of you do this, but it got me thinking about “help” systems in general. Why are we still using these static help files? I can see where they are useful as fallbacks, but in this connected world, what about leveraging incremental updates and indexing engines such as Google? Could a filter alogrithm be created that streams down the links to your desktop? The value added would be in the additional processing applied to Google. Then again, maybe it’s not really necessary. That’s the beauty of Google, I guess. Hmm.
What does this say about how these indexes should be managed in the future?
First and foremost it suggests that in the Google world strong communities of developers that share information are a big advantage. Together they go far beyond the value of the built in help.
What you are suggesting is very similar to Gurunet or Babylon – instant, context-sensitive information. The added touch is the fact that information is “pushed” to you in an ambient way (mouseover for function details).
Obviously, the more specific the application, the more relevant and useful the indexing/linking can be.