Bottleneck

Friday, 13 April 2007

Daniel Jalkut has a nice piece on Apple’s “iPhone development effort caused Leopard to fall behind schedule” excuse:

The best we can hope for is that it is only sleazy marketing bullshit. Because if what Apple’s telling us is true, then they’ve confessed something tragic: they’re incapable of building more than one amazing product at a time. The iPhone looks like it will be an amazing product, but if Apple can’t keep an OS team focused and operational at the same time as they keep a cell phone team hacking away, then the company is destined for extremely rough waters as it attempts to expand the scope of its product line.

No argument that this is bad news for Apple. The original ship date of June was more than two years after the release of 10.4.0, which would have already made it the longest stretch between major releases of Mac OS X. But it’s unfair to make an “Apple can’t walk and chew gum at the same time” argument regarding this delay. Leopard and iPhone aren’t two wholly separate products. They’re two products built on the exact same foundation: OS X.

Compare and contrast with the original iPod, which was a wholly separate project from Mac OS X. The original iPod was released in October 2001, and developed during the preceding six months. That’s right in the middle of the most productive stretch in Mac OS X history — 10.0.0 was released in March 2001, 10.1.0 was released in September, and 10.2.0 was released less than a year after that in August 2002.

Jalkut continues, again expressing what I suspect many people are thinking:

What happens when the phone takes off, and Apple’s stuck following through on their Mac OS X commitments? “Sorry, no iPhone 2.0 until 2009 - we’ve had to move everybody back to OS X!” …

Perhaps they’ve regained success too quickly. If a company with a market capitalization of $80 billion, and a cash account of at least $6 billion, cannot hire enough people to build three of the hottest, most demanded products in consumer electronics (the iPod, the Mac, and the iPhone, if it’s not obvious), then maybe it’s time to reevaluate their modus operandi for attracting and retaining talent.

The problem isn’t that Apple hasn’t grown enough in the last five years. (There’s a reason they’re building a new campus in Cupertino.) The problem is where they’ve allocated the talent. According to an engineering source at Apple, the OS group headed by Bertrand Serlet has only grown “just under 10 percent” in the last five years. The apps group, on the other hand, has grown significantly during the same period, and now outnumbers the OS group.

There’s a difference between throwing new hires at a late project (which almost never works, and almost always in fact makes things worse), and allocating the OS team the resources it needs. OS X is being asked to do far more — powering both iPhone and Apple TV while continuing its role as a desktop and server OS for the Mac — but with almost no additional engineering talent.

Leopard is far too big a project for any simple explanation to fully explain why it has fallen behind schedule. But from what my sources at Apple tell me, this is a case where the PR spin is in fact essentially true. It’s not so much that talent was pulled from OS X to work on the iPhone, but that they were pulled from the Mac part of OS X to work on the iPhone part of OS X.

There’s still going to be a major new release of OS X in June. It’s just not the one we’d been expecting.