gavinshearer.com

« Meet And Greet | Main | Another Childhood Memory Bites The Dust »

January 22, 2006

Prediction: XP Is The "New Classic" Under 10.5

Since the unveiling of the Intel Macs, the Mac blogosphere has been all atwitter about the possibility that the new iMac and MacBookPro (it's two weeks later, and I still hate that name) will run Windows.

Most of the conversations are focusing on the technical details of such a move, looking at whether or not Windows supports Apple's new EFI BIOS (XP does not, but Vista will), and how hacks might conceivably allow one to get around any technical challenges. But the envisioned scenario - installing some kind of shim to dual-boot between Windows and OS X - requires a fabulous level of geek knowledge in order to work. While the Alpha Nerds among us will certainly make this happen (just as they've managed to poke and prod OS X for Intel so it runs on, say, a Dell), Ma and Pa Mac User will certainly never go to these lengths in their day-to-day lives.

Most of these conversations, I think, are missing the point. There is absolutely no doubt that the new Macs will run Windows. In fact, I predict that Apple is working on a system to make this happen, and it will ship with the next iteration of Mac OS X - 10.5, or "Leopard".

(Quick disclaimer: Yes, I work for Microsoft. But, no, I don't know anything about any of this from my job. This is just a best guess, based on my having watched the Mac market for a long, long time.)

Apple's decision to select Intel processors for its machines has three big benefits. Of these, Apple has only talked about one of them. They are:

Benefit #1: Performance Per Watt. Yes, Steve got up at WWDC in June and waxed rhapsodic about Intel's "performance per watt" - for each X of juice the chip takes in, it pumps out Y amount of performance. The thinking here is that the Intel iron allows Apple to provide G5-class performance with substantially less energy (and, thus, heat) than the G5 itself. Theoretically, this will allow Macs to be smaller, offer better battery life, and sport more innovative designs (big, chunky heat sinks tend to be a design constraint, you know?).

I expect some new, kick-ass designs from Apple later this year.

Benefit #2: Reduced Point Of Difference With The PC World. A secondary, but little-discussed benefit of going to Intel is that Apple no longer has to shoulder the burden of proof that the PowerPC is a good processor. Apple has spent hundreds of millions of marketing dollars working to convince people that the PowerPC chip gave Macs a performance edge. This was, ultimately, a losing battle. Consumers only have so much attention to give any one company, so every precious moment Apple used to defend the PowerPC was one less moment it had to talk about what makes the Mac experience better. By and large, people don't care about processors - they care about buying something that works as advertised. And for a lot of people, something that's too different - namely, a different processor plus a different operating system - feels risky, which then makes them averse to buying your stuff. Most people feel safe in numbers, so they tend to go with what "the crowd" goes with (e.g., Dell). It's the "When Harry Met Sally" principle of computer buying: "I'll have what she's having."

Apple no longer has to fight this fight. Instead, its message can be 100% focused on what makes the Mac a better computer. The debate is now about features, benefits, and ease-of-use. I think Apple is well-suited to fight that battle.

Benefit #3: Binary Compatibility With Windows. This third point is at the heart of this post. By going to Intel (as opposed to the Sparc, or the Cell, or any one of a billion other chip designs Apple could have selected), Apple now has binary compatibility with Windows, and, by extension, the entire family of Windows applications that exist in the world. This is the centerpiece of Apple's strategy for the next 5 years. Leopard's going to exploit this benefit like nobody's business.

Look, it's been possible to run Windows on the Mac for at least the last decade. Programs like Virtual PC have allowed you to install and run Windows as a Macintosh application. The big drawbacks of the Virtual PC approach were performance and experience. Performance-wise, since Windows is compiled for the Intel processor, Virtual PC had to translate the Intel instruction into a PowerPC instruction (much as Apple's Rosetta technology does, but in the reverse direction). This translation process takes time, and makes applications feel sluggish. And on the experience side, all your Windows programs run inside your single Virtual PC window, which means you have to drag-and-drop your files and whatnot between your "real" Mac and your "virtual" Windows machine. This is not optimal for lots of reasons, but the Virtual PC team did an amazing job of making it as seamless as possible.

Today, all that's out the window.

With binary compatibility, we don't need to worry about emulation any more - it's Intel all the way down. Thus, a Windows application running on an Intel Mac will be as fast as it would be running on the same-speed machine from Dell.

And on the experience side, well, I suspect that Apple is working to make that pretty seamless. And the only way they can do that is to build compatibility for Windows software right in to the operating system directly. As it happens, Apple has a lot of experience doing this - namely, with Classic and with X Windows. In both cases, Apple made it possible for programs written for other operating systems and environments to work natively, and seamlessly, on Mac OS X.

"But why?" you ask. "Why would Apple want to embed Windows compatibility into Leopard? Isn't that just sleeping with the enemy and helping Microsoft?"

First, as far as "sleeping with the enemy," let's be clear. Apple and Microsoft have a deep and complex relationship. Microsoft develops some great software for the Mac, and Apple cross-licenses patents with Microsoft. Steve Jobs and Bill Gates have known each other for decades, and built this industry together. Yes, Apple and Microsoft compete in a lot of areas, such as operating systems, digital music, and video standards, but the technology business is one where everything's in a shade of gray, rather than black-and-white. So if Steve's strategy to grow the Mac business makes Bill a few million in new Windows licenses, well, Steve'll do it. No question.

But as to "why?" - Apple wants to build binary compatibility into the operating system because it's necessary. Getting consumers to switch to a Mac is much, much harder than anyone at Apple originally thought it would be. The "Switch" campaign of a few years ago is regarded as a failure not because the message was wrong, or the ideas were bad - "Switch" failed because nobody at Apple really understood what they were asking consumers to do.

Your PC is a crazy thing, a thicket of programs and data files that you rely on to get your work done. And, generally, every program that exits for Windows exists for the Macintosh. But while "generally" existing is good enough to keep the platform viable, it's not enough to get people to jump ship. People get used to one specific thing, and the "general" replacement just doesn't cut it.

Here's an example. Let's say your grandmother is sick of spyware and all the cruft of her PC. So you say, "Hey, Grandma, let's get you a Mac." She does e-mail, of course, writes the occasional letter, balances her checkbook and browses the Web - should be no problem. The Mac does all that stuff.

But as you get in to talking with her, you then find out that she also uses her computer to play solitaire, and has developed a real love of Sudoku puzzles. And furthermore, she loves the programs she's using to play these games on her PC today. Suddenly, your "simple" migration to the Mac is now a hairy pain in the ass - your grandma doesn't want the Mac if she has to give up her favorite solitaire game. And so, after much hemming and hawing and thinking about it, you buy her a new HP laptop and double up on your antivirus subscription. No switch for you.

Trust me: this scenario has played out millions of times over the past few years. There's always that one damn application that the potential switcher loves, or cannot live without. And that's what's been killing Apple's potential for growing the Mac base in any significant numbers.

By including Windows compatibility in Leopard - making it, in effect, the "new Classic" - Apple can now sell a machine to your grandmother. She'll use Safari for the Web, Mail.app for her e-mail, and will run Windows (as a process) for her Solitaire game ... until she finds a new, better Solitaire game for the Mac, and then kicks the old Windows one to the curb. (And, I should point out, that Windows on a state-of-the-art Mac will feel smokin' next to a couple-year-old PC.)

Apple's gonna migrate Windows users to Mac OS X the exact same way they migrated Mac OS 9 users to Mac OS X: through embrace and extend. Take the old OS, run it as a process, make it seamless. Then sell users on the beauty and power of native Mac applications.

Legally, I predict that Apple will get Windows on the Mac the same way Connectix did with Virtual PC - they'll just license Windows from Microsoft. This has the advantage of providing a higher-quality Windows experience, and also ensures that Microsoft is making more money (and, hey, we like to make money) on the Mac.

In terms of migration, Apple will probably offer something like their migration assistant to make it dead-bang-easy for people to move their files from a PC to the Mac. I envision some kind of bootable CD that turns your old PC into a big USB Mass Storage device. From there, you'll simply run a USB 2.0 cable between your Mac and PC; the Mac'll suck over the entire drive, creating a disk image in the process. The disk image will be a freeze-dried copy of your PC data, files, and whatnot; this prevents any kind of nasty commingling of files in the Mac OS X file system, and also prevents disk formatting problems.

If Apple's really sharp, they'll take this migration one step farther, ala Move2Mac, and import your IE favorites, convert your Outlook Express mail to Mail.app, and even set up users on your Mac that are similar to those on your PC. (They might even be able to move your Windows license, too, which could save you some money.)

So yes, Virginia, you'll be able to run Windows on the Mac. And soon, I'd wager - Leopard is due in "early 2007" (which sounds like a MacWorld keynote to me). And from there, I predict that Apple will see their market share double in about two years. The barriers to switching will be, for all intents and purposes, gone.

Interestingly, the new Intel macs are the first Mac OS X machines not to run Classic. In its place, Apple is going to introduce the "new Classic" - Windows XP. It's funny how things work out.

Posted by Gavin Shearer at January 22, 2006 4:52 PM. Posted to Apple.

Comments

Posted by: Khan Klatt at January 22, 2006 11:25 PM

Posted by: Chris at January 23, 2006 8:49 AM

Posted by: gr33n at January 24, 2006 3:58 AM

Posted by: coelomic at April 7, 2006 12:57 PM

Posted by: Anonymous at April 7, 2006 1:07 PM

Posted by: The Valrus at April 8, 2006 7:28 AM

Posted by: Drew Thaler at April 12, 2006 8:37 AM

Posted by: Drew Thaler at April 12, 2006 8:58 AM

Posted by: Christopher at April 30, 2006 2:35 PM

Post a comment

« Meet And Greet | Main | Another Childhood Memory Bites The Dust »