The new frontier

Anyone hoping that Steve Jobs would make some sort of bombshell announcement during his keynote at the 2007 Worldwide Developers Conference came away disappointed.

The news about the upcoming Mac OS X ( Leopard ) was mostly a refinement and clarification of what Apple outlined last August at WWDC 2006. The new things Jobs did announce—the translucent menu bar; the Dock’s pseudo-3-D tray; the revised, iTunes-like Finder; and the consistent systemwide look-and-feel of windows— were mostly cosmetic.

But this isn’t necessarily bad. The relatively low number of significant changes to Leopard since WWDC 2006 is, more than anything else, a sign of OS X’s maturity.

Solitary refinements

From the late 1980s until the release of OS X, Mac users obsessively speculated about the next Mac OS. Pink, Taligent, Copland, Gershwin—the list of next-generation Apple operating systems that never came to fruition is long. Users wanted something new, something that would be as big a leap from the then-current operating system as the original Mac was from the Apple II. When Apple transitioned from OS 9 to OS X, we finally got it.

The initial major releases of OS X were exciting because the early versions had such gaping holes and serious performance flaws. The only reason it was possible for Apple to improve OS X so much between 10.0 and 10.4 was that there was so much room for improvement.

I’m not arguing that there’s no room left. It’s just that OS X 10.4 is so fundamentally good that future upgrades are likely to be on the scale of small refinements.

Apple’s long-term strategy for desktop computing seems to be refining OS X, not replacing it.

This history is analogous to that of the automobile industry. In its early years, the state of the art advanced at a remarkable clip. Today, new cars come out each year, but with small refinements. Those changes add up: a 1997 car (even in mint condition) is clearly distinguishable from a 2007 model. But a 2005 and a 2008? Not so much. That’s pretty much where we are with OS X. Tiger is the 2005 model; Leopard is the 2008.

The 3.5-inch revolution

With the iPhone, on the other hand, Apple is heading into uncharted territory. The fundamental elements of the Mac interface are overlapping windows, the menu bar, and the mouse pointer; the iPhone has none of these. Every Mac ever sold has had both a keyboard and a pointing device providing single-pixel precision; the iPhone has no physical keyboard, and while using your finger is convenient, it’s far less precise than using a mouse, trackpad, or stylus.

To accommodate this lack of precision, the iPhone provides much bigger on-screen targets than any Mac, on a smaller screen. As the size of desktop displays has increased, the Mac user interface has been able to display more and more information on screen. The iPhone deliberately displays less. When listing messages or songs, for example, rather than trying to fit as many as it can on screen, the iPhone uses a generous row height, making them easier to tap.

The iPhone interface isn’t about refining and improving something that already exists. It’s about completely new ideas in user-interface design. So it’s no wonder that many Mac developers so desperately hope to write their own software for it.

Currently, Apple’s only statement about third-party software development for the iPhone has been that the iPhone will run Web applications through its built-in Safari browser. That’s a great feature for such a tiny device—no argument there. But a Web page inside a browser is no more a venue for a real iPhone app than it would be for a real Mac app.

Even if Apple privately plans to allow the development of third-party iPhone software in the future—with an iPhone variation of Cocoa, Dashboard-like widgets, or both—the company won’t say anything until those plans are ready. “Underpromise and overdeliver” is a strategy that has served Apple well in recent years.

There has been some speculation that Apple might bring iPhone innovations, such as the multitouch screen interface, to the Mac. Anything is possible, but I think that multitouch Mac displays are unlikely and would be unwise. And the simple truth is that OS X doesn’t need an interface revolution.

The iPhone’s screen measures just 3.5 inches, but it’s now the biggest frontier in interface design.

[ John Gruber writes and publishes the Mac Web site Daring Fireball. ]

Subscribe to the Best of Macworld Newsletter

Comments