Dr. Dobb's Journal November 2006
A YEAR AFTER ANNOUNCING the planned move to Intel processors, Apple has completed the transition—there's not a nonIntel Mac left in the lineup. Given the performance versus price of its latest Pro model, Apple has arguably also repealed the coolness tax that its hardware has always carried. And not only can you run Windows software natively on all new Macs today, but there are three competing technologies for running Windows. Attendees at Apple's Worldwide Developer Conference saw new tools and a preview of the next version of the operating system, and heard tantalizing hints that this was only the tip of the iceberg and that there was much more to come. It's tempting to do a little guessing...
But consider poor Jim Carlton.
It's 1997, Apple's blackest year, and Carleton has just finished the final interview—with Steve Jobs, no less—for his first book, Apple (Times Books, 1997). After many such interviews, many months of painstaking research and fact-checking and fussing to make sure he had it all just right, he finally turns in the manuscript, and waits.
The flow and flux of events, however, does not wait for publishing schedules. Within weeks, Apple has completed a pundit-flummoxing profitable quarter. And it's no fluke: Every quarter that year turns out to be profitable. The stock starts looking healthy, the company has money in the bank, the product line looks solid—the turnaround is complete.
And here's how Carlton ended his book, a book that came out after it began to be clear that Apple was coming back: "To avert the seemingly inevitable loss of its independence, Apple, in short, needs more than a few breaks. It needs a miracle." The last chapter of the book reads like an epitaph. The book was, in part, out of date when it hit the racks.
The moral of the story is: Big companies collect a lot of barnacles as they grow, and become hard to turn. But even a big company can change direction quickly in certain circumstances. In times of trouble, facing bankruptcy or acquisition, a company can decide to make ruthless cuts. Apple did that, and while turnaround man Gil Amelio certainly gets a lot of credit for this, it should also be noted that the current CEO is more willing than most to make ruthless cuts where he doesn't think a person or project measures up. His name has even become an eponym, as in, "The planet Pluto just got Steved."
What distinguishes Jobs from a turnaround man like Amelio is that Jobs is perfectly willing to make radical changes when the company isn't in trouble. And so Apple keeps evolving, and keeps us guessing.
It seems as though it took Apple exactly a year to transition its entire computer line from PowerPC to Intel. It seems this way because Steve Jobs made the initial announcement in his keynote at 2005's Worldwide Developer Conference. In this year's keynote, he announced the final Steving of the PowerPC Mac with the release of the Intel Xeon "Woodcrest"-based Mac Pro. But WWDC 2005 was in May and WWDC 2006 was in August, so it was actually 15 months. In terms of the OS, of course, there essentially was no transition: OS X has been built for both PPC and Intel (sort of secretly) since Day One.
For third-party developers, moving the software to Intel was a different story.
It's interesting to note that the last time Apple did this, when it moved from the Motorola 68000 chip family to the IBM/Motorola PowerPC family, MetroWerks saved Apple's bacon, so to speak, by providing in Code Warrior the development environment developers needed to create those third-party apps. It's interesting because this time around, MetroWerks is out in the cold—not Steved, exactly, because they did it to themselves, but out of the loop. The upshot for developers, though, is that it's Apple's Xcode or—well, it's Xcode.
Actually, it needs to be mentioned that Intel has been providing some key tools for development, and has been focused on integrating them into Xcode, as Intel's James Reinders told us earlier this year (www.ddj.com/184406371).
A year (15 months, but who's counting) after the announcement, a surprising number of apps have been moved to the Universal Binary format that, like the Fat Binary of another era, bundles two versions of the app in one file and spares user pain. Only slightly painful for most users is Rosetta, the technology that lets unmodified PowerPC apps run on Intel Macs, albeit slowly. But for Apple's third-party developers, the transition has been, generally, as easy as Apple predicted.
Although most third-party developers seem to be moving rapidly to Intel, two big players are playing by their own rules.
Microsoft's Macintosh Business Unit is hard at work making the next version of Office for Intel Macs everything you expect Microsoft Office to be, so long as you don't expect it to be capable of running VBScript scripts. Ain't gonna happen. And Virtual PC: That's a deader. Too bad. It would have been fun to compare it with all the competitors that have sprung up.
Then there's Adobe. Adobe says very positive things about wanting to do it right and about how its next release, which will be delivered on Adobe's schedule, not Apple's, will fully support Intel-based Macs and we will all be very happy. In the meantime, Adobe advises not buying an Intel Mac if your top priority is Photoshop performance. Apple can't be thrilled to have a major third-party vendor discouraging people from buying Apple products. And one notes, from trolling the places where such people hang out, that Photoshop users who have tried running Photoshop under Rosetta mostly find it not that terrible—although, of course, these would be people who just upgraded their hardware, so there are likely to be offsetting factors helping out with performance.
There's also a bit of sniping, at the Adobe engineer blog level, about Xcode not being up to the job. But when Adobe talks about needing to move to Xcode only this year, you wonder: Didn't they get the memo about how important it was to move to Xcode?
Ah, well. Reports are that Windows Photoshop runs really well on the new Intel Macs using Boot Camp.
Windows on a Mac, right. Companies spend millions to get their cute little phrases and product names drilled into the public consciousness, but in 2006, the words "on a" with a noun at either end, instantly conjures up the image of Samuel L. Jackson dissin' vipers in a B movie.
If anything about the latest Macs compares with "Snakes on a Plane" for newsworthiness and weird-soundingness, it's running Windows applications on Apple hardware. It seems to have taken Apple roughly that same 15-month year to bow to the inevitability of this and get out in front of it, sort of, by giving all Intel Mac users this capability in the form of its dual-boot Boot Camp software, to be bundled with the next version of OS X.
But Boot Camp is not the only way to run Windows on a Mac, nor, for some purposes, the best. Parallels (www.parallels.com) lets you run Windows apps just like another app in the dock. And VMWare (www.vmware.com) gives them some competition. As one observer put it, nobody's going to miss Virtual PC.
Last year at WWDC, we were developer-statused, briefed, and NDA'd, but this year we opted to be ignorant journalists. Everything reported here has already been made public, though not always with Apple's approval. ("Leopard Gets BitTorrented; Apple Employees Get Steved.")
One anonymous developer said that what developers saw was "a lot cooler" than what was shown in the keynote walkthrough of the new OS X 10.5 ("Leopard") features, and predicted that people using Macs in networked business settings would be particularly pleased.
What else looks interesting? I jumped on Dashboard when it was released last year, but haven't used a widget in months. The new DashCode development tool for creating Dashboard widgets could get me interested again, that and the end-user Web Clip tool that turns parts of web sites into widgets.
The 64-bit support is a big deal, and it runs through Cocoa and Carbon on PPC and Intel, allowing 32-bit and 64-bit apps to run natively side-by-side.
Time Machine is an impressively realized backup tool that will interest developers for its APIs and as a demo of Core Animation. One writer suggested that this animation technology could lead to UI animation craziness reminiscent of the font insanity in the early days of desktop publishing.
Apple's Mail, iChat, and Spotlight search apps have all been upgraded, and again for developers, the interest may be the ability to tap into these application's new functionality in their own apps. The Text Engine provides system-wide text-processing enhancements like smart quotes and discontiguous selection. The promise of a resolution-independent UI is in there, too.
XCode 3.0 was announced at WWDC, and ships with Leopard next year. The consensus seems to be that this is a major improvement. This is good, because you don't have much choice if you're developing for Mac. One new tool is Xray, an app performance optimizing tool with some interesting features. Ruby on Rails, Subversion, and Apache 2.0 are expected to ship with 3.0.
For a moment there, it looked like Apple was retreating from open source, but the launch of the Mac OS Forge community development site and the opening of the source for the Intel 10.4.7 kernel and other tools (Bonjour, launched) should reassure the doubting.
WebObjects will probably join that collection of open-source technologies. Apple uses WebObjects extensively in its own web apps, but has had little luck selling it, and doesn't even seem to be trying any longer. Rumor is that it'll go open source, except for patented bits, soon after Leopard ships.
What's been thrown out of the sleigh?
Well, Leopard requires at least a G4 Mac. So the G3, to which Apple owes so much, has been Steved, as has Classic mode. HyperCard stackheads can take consolation, though, in the fact that there are at least three emulators that run on Intel machines and support Classic apps: SheepShaver, Basilisk II, and Mini vMac. The last of these actually emulates a Mac Plus!
Languishing, but not actually discontinued, is Apple's .Mac online service. From the way Apple talks about it, you'd think that .Mac was competitive with other such offerings. And Apple has even come up with the ad slogan "share" to promote .Mac. One is tempted to speculate about what Apple might have in mind.
But that would be guessing.