Is the iPhone 3G Defective?

Truth be told, I've been writing a lot about the iPhone 3G lately because, despite the tone of my last entry, I'm still thinking about getting one. I need to replace the Q as early as next month, and, given my compendium of requirements, the iPhone remains at the top of the list. As I've noted before, the walled-garden approach isn't unappealing to me; I'm a computer-science major, but I have a business to run and a job to do. So the Mac is a decent platform to base it all on, and the iPhone goes hand-in-hand with the Mac. Is that great marketing, or what? And I'd be amazed if I ever downloaded more than two apps for the iPhone; I just need the basics for business use, and at the core that's a desktop-class browser. So I should be all set.

But there have been a number of disturbing articles about the iPhone 3G and related functionality in the press since the device was announced. I wrote earlier about Walt Mossberg being less than enthusiastic about MobileMe, which would be critical in my use of the iPhone, and this article in BusinessWeek seems to confirm that heads did indeed roll over this episode. It also appears that the problems are getting fixed. No matter; I rarely buy Release 1.0 of anything, especially for production use. Apple isn't the only company with quality problems; indeed, they are among the better vendors out there.

Of greater interest, though, is this piece at MarketWatch.com, which talks about a possible hardware flaw in the iPhone 3G. Some users have been complaining that throughput is slow, calls are dropped, and that fundamental issues with the Infineon chipset used in the iPhone 3G are to blame. That's a real possibility; designing wireless chips is among the most difficult engineering tasks on this planet, and all chipsets have limitations and perhaps even flaws. Even the way the chips are arranged on the circuit board can affect performance. But given just the statistical nature of radio itself, and well, variable performance has been a core element of the history of wireless since the beginning. You want perfection? Use wire. You want convenience? A little inconvenience is unfortunately part of that experience, and likely always will be. Can the vendors do a better job in engineering? Sure, but that's why the next product is (almost) always better than the one we use today.

Vendors can regardless do more, as I've written elsewhere, to improve the fundamental quality of their products (as Apple realized in the MobileMe case), and quality will ultimately be a big differentiator in the market. Carriers should properly set the expectations of users, and not just market numbers that represent the performance that a given product will never, ever exceed. The best way to create a satisfied customer is to set their expectations properly, and the cellular industry as a whole, whether we're talking throughput, coverage, or whatever, just isn't very good at this today.