As usual, Steve Jobs' keynote presentation at the WWDC in San Francisco did not disappoint. Most of the announcements focused on the next major release of the Mac OS X operating system, named Tiger (10.4). If you have an hour watch the demos -- they are best seen, not read about.

Out of all of the stuff announced for Tiger, the searching API named Spotlight will have the most impact on users. It is quite interesting that not long after (in software development terms) Microsoft announced its intentions for Longhorn that the advanced searching features have appeared in Mac OS X.

Can we fairly attribute this to Microsoft's openness about Longhorn as a competitive disadvantage? I think partially yes, but we could also argue that given the mainstream success of Internet search engines that better searching technology on the desktop would become the next logical step. Using this reasoning one could argue Microsoft just telegraphed an obvious direction and it won't hurt them.

Nevertheless, Apple is goading Microsoft at the WWDC, which Steve says is a little "friendly competition". Best not get too "friendly" with the developer of the most popular office suite, eh Steve? They may bite back. The tone of the posters is a little over-the-top I think, instead of Apple's usual and preferable passive aggressiveness.

Who knows, Apple could have already been heading in this direction without Longhorn's inspiration, but Steve sure feels good about beating Microsoft to the punch by at least a year.

So what is Spotlight then? It's an API for searching files based on content. Indexes, which are maintained at a level just above the file system presumably, are used to retrieve files on the system matching search terms. The demo found files pretty quickly in a system with a good number of files in it.

Indexes, which Google and other Internet search engines also use to match search words with web pages, are created by taking the words of a document and pointing them to that document. Do this for every document in the file system, combine them and now you have an easy way to find documents containing a given word.

This is pretty good for file containing words (documents) but binary files are a bit different. The indexer has to be able to understand the file format in order to extract meaningful information about the file from it. This information about the contents of a file is commonly called metadata.

The most frustrating part about metadata is that users are often too lazy to input all of it. Possibly with better searching tools like Spotlight users will be motivated to input good metadata so that files can be found easier. It's a chicken and egg problem: users right now don't see much of a use for it.

As Steve mentioned in the keynote, searching will reduce the need for defined directory hierarchies for organization of hard drives. Today people spend a lot of time sorting files into their correct directories so that they can be found later. But with Spotlight people will save time by not having to organize their hard drives and doing explicit searches for files.

In the future it's not unreasonable to envision a dynamic file system hierarchy -- one that is created on the fly with metadata instead of rigidly in a file system. The user is then free to browse this dynamic file system presented to them based on criteria they choose. Like "show me all of my pictures, sorted by date and location".

What does this mean for AudioMan? In the operating system of the future metadata will be more important than it is now. We'll need tools to make metadata "tagging" of files more efficient and easy, and AudioMan intends to do just that. Don't be surprised if iTunes starts taking metadata tagging efficiency more seriously after Tiger is released.