In recent years, Mac aficionados have been saddled with a difficult burden.
Convinced that their platform is easier to install, use, maintain, upgrade, and support than any other desktop OS available, they've been forced to defend the Mac OS against a market that seems to have inexplicably, all-but-decisively gone the other direction.
Try sitting a Windows user in front of a Mac, sometime. Tell them how many years Apple spent crafting its user interface.
Explain how elegantly the Apple hardware is designed, how easily the parts and peripherals interconnect. Explain the sophistication of the machine's supercomputer-class G4 processor. Nine times out of ten, I guarantee you'll get the same question in response.
They'll want to know why everything grinds to a halt when they hold down the mouse button.
Enter Mac OS X, the ultimate product of the turbulent last few years of Apple's history. The X is really a Roman numeral 10, which might have you believe it's just one generation away from the current Mac OS. In truth, this new OS is another beast entirely.
"Mac OS X features the battle-tested Mach 3.0 kernel," trumpets Apple's Web site. To most Windows users, those words probably seem no more significant than Apple's decision to ship its slow, crash-prone Macs in a rainbow of colors. To the much-maligned cabal of Apple-loyal techies, however, they may just mean salvation.
A kernel is an extremely low-level piece of software that runs at the core of an OS. Think of it as the grain of sand at the center of a pearl; all other software is layered on top.
The kernel operates constantly, doling out resources like memory, hard drive space, and screen real estate to every program running. Each system task in turn defers to its control. Thus, if it's well designed, any number of programs could potentially run at once, without even being aware of one another's presence.
The kernel that powers Microsoft's recently released Windows ME isn't much changed from the one the Redmond software giant originally unveiled in 1995. But lined up side by side with Mac OS 9, there's literally no contest. You can't pit the venerable Windows kernel against that of the Mac OS — because the Mac OS doesn't even have one.
In the early 90s, while Microsoft was working to introduce modern OS concepts into Windows, Apple had become mired in its now-legendary arrogance. Though new versions of the Mac OS had appeared, functionally they were little improved from the MultiFinder task-switching software introduced years earlier.
Without an omnipotent kernel to mind the store, applications running on the Mac OS were free to run wild with system resources, trashing drives and writing to memory where they didn't belong.
Worse, multitasking relied on every application being written such that it would cooperate with all the others. Any program could wrest control of the entire CPU at any moment; the OS merely trusted them not to. Case in point, when the user held down the mouse button — one showstopper that's never been remedied.
At the system level of the Mac OS, the lunatics have been running the asylum. The goal of Mac OS X isn't just to introduce new technology. It's time to restore order.
Job #1 in curing the Mac's ills was getting a kernel in there to take the reins. But as it turned out, this task was no easier than Apple's search for a CEO to run its business affairs.
One of the earliest efforts at a kernel for the Mac OS was an attempt to write one brand new, from scratch. Code-named Copland, this undertaking promised pre-emptive multitasking, protected memory, and all the other modern OS buzzwords that Windows 95 users obliviously enjoyed.
But the project became mired in the delays and infighting typical of Apple at that time. Its release date came and went, and eventually it was scrapped completely. Copland is now relegated to that Hall of Fame reserved for all legendary Apple bungles, nothing more than over-hyped vaporware.
Meanwhile, Microsoft had left Apple for dead in the OS game. Windows 95 had swept the desktop market, the Mac's market share plummeted, and Redmond moved on in search of bigger game.
Their next product, Windows NT 4.0, had a tougher and more sophisticated kernel than ever before. This time, their intended market wasn't home users. They were looking to push Windows into the world of enterprise computing, where UNIX-based servers had traditionally done the heavy lifting.
Some Mac loyalists see it as fitting, then, that when Apple finally introduced its next-generation OS, this time it would sport a time-tested UNIX core: the Mach kernel, developed at Carnegie-Mellon University. And this time it wouldn't be reserved merely for the business server market. Mac OS X aims to bring UNIX to the desktop.
Along with it comes everything the Macintosh market has been waiting for. True multitasking, managed by the kernel. Protected memory, so no one application can knock out all the rest. Efficient networking. And the entire heritage that UNIX has to offer.
Still, it's a difficult road, and plenty of hard work, that lie ahead. Few who have received the $30 Mac OS X Public Beta from Apple can claim it meets all their computing needs.
It supports few printers, and can't run Apple's AirPort wireless networking cards. Many standard add-on storage devices, like SCSI cards or PCMCIA hard drives, won't let it boot. And without standard UNIX compilers and development tools, casual users won't be able to add new components to it, as they would a Linux system.
The biggest challenge to UNIX's acceptance amongst desktop users, too, has traditionally been the user interface. UNIX-based systems tend to feel clunky to those accustomed to the flash and fuzz of a Windows or Mac OS box.
Apple has made tremendous strides in this area, and Mac OS X's translucent interface is beautiful to behold. But some aspects of the design still betray their heritage in OpenStep, the UNIX-based OS Apple bought from Steve Jobs' NeXT.
The Mac OS X Public Beta is no Cadillac. It's a hot rod. The hood is open, and it still needs a muffler.
But for all those Mac stalwarts who have waited so long for proof positive that their platform really is better, it serves as a small glimpse into a comforting future — and for that, $30 is a small price to pay.
