The video demos of new features in OS X Leopard are pretty chill — OK, more than chill — some of them are downright amazing. But I’m trying to wrap my head around the release of Safari for Windows.

With iTunes for Windows, it was a slam dunk – you can’t sell iPods and tracks to people who can’t reach your platform. But with Safari, it’s not so clear cut. What are they selling? Ostensibly, it’s about giving Windows developers access to the browser that will be running on the iPhone. But I’m not buying that that’s the whole reason. Developers are just too small an audience to warrant the work it must have taken to do the port, and to support it going forward.

There’s the old “gateway drug” argument – give Windows users enough tastes of Mac elegance – and in this case a faster browser than anything available on Windows right now (Apple claims Safari 3 is twice as fast as Internet Explorer 7 on Windows, and 1.6 times faster than Firefox 2) – and eventually they’ll wander over to take a closer look at the whole enchilada*. But how many Windows users are going to care? Those who care enough about security and extensibility to try another browser are already using FireFox, and Safari doesn’t have FF’s thriving plugin landscape going for it. Speed alone isn’t going to cut it.

So… they’re going to end up with a tiny percentage of developers and geeks running Safari on Windows. And this benefits Apple how? Maybe I’m wrong – maybe the need to provide a platform for Windows iPhone developers is reason enough, but somehow that doesn’t ring true. I think there’s another shoe ready to drop, lurking stage left.

* Update: I wrote that bit about “elegance” before seeing any reviews of Safari/Win after it was released into the wild. Now that the opinions are starting to roll in, I think it’s safe to say that this beta was released long before it should have been. By all accounts, Safari/Win so far appears to be a steaming pile of $%$%!@ with little to recommend it.

In January 2003, I predicted that major browsers would have RSS readers built into them within six months, which would have put the first release of such a feature at around July 2003. Obviously, that didn’t happen. At WWDC today, Apple unveiled the major features in Tiger (OS X 10.4). Among them is an integrated RSS reader for Safari. That puts my prediction off by a year. Except that Tiger won’t be out until 2005, so make that 18 months. My crystal ball must have been hit by a stray neutrino.

A few readers have written me over the past week speculating about whether HFS+Finder in Tiger would more closely resemble BeOS’ spectacular BFS+Tracker combination, given that Dominic and Pavel have been working at Apple for the past year and a half. The answer is… sort of. Spotlight is going after metadata in a big way, and is making system-wide, instantaneous search on any type of heterogenous data seamless. The “Keywords” feature may or may not be similar to BFS’ Attributes. The key to making Spotlight as fast and flexible as BFS+Tracker will be, for me, whether attributes, er, keywords, are 100% customizable into arbitrary metadata forms, whether the metadata indexing is fast, automatic and efficient, and whether Apple finally releases a complete cousin to Be’s incredibly powerful FileTypes panel.

Also cool in Tiger: The very slick Dashboard (did Apple purchase Konfabulator?), full videoconferencing in iChat, and a powerful-looking scripting front-end called Automator. Oh, I wouldn’t mind a 30-inch aluminum-encrusted display, either.

Update: The Register confirms that Apple didn’t buy Konfabulator – they pulled a Watson on it (thanks Michael). I got a kick out of one of the propaganda posters Apple apparently has out at the WWDC: “Redmond, start your photocopiers.” Which is especially ironic given all the copying that Cupertino is apparently doing.

It’s funny – an Apple-branded web browser is the last thing I thought I wanted out of Expo, but I’m already head over heels for Safari. Earlier this morning I was wondering why they would have chosen the lesser-known KHTML rendering engine over Gecko, but after feeling its speed and reading that the codebase is less than 1/10th the size of Gecko’s, I get it.

Safari’s bookmarks implementation is a thing of beauty. I’ve wondered since the mid-90s why no one ever seemed to get bookmark management right, but think Apple has finally cracked the egg. The one aspect of bookmark management I miss from BeOS is the ability to add keywords to bookmarks and then find similar bookmarks via live keyword queries. But if you make sure you give good descriptive titles to your bookmarks, the existing Find function works just fine.

They copied all the hotkeys over from Explorer, so everything works as expected. Even the almighty Cmd-click to open a link in a new window.

I’m always surprised to see how many people gripe about the brushed aluminum look. Personally, I’m sick to death of looking at white stripes and would be happy if every app on my system went brushed aluminum. However, it does seem like Apple has violated their own guideline to use brushed aluminum for apps that replace real-world devices. What device does the browser replace?