In August, a programmer named Alex Sokirynsky wrote a clever app to let iPhone users stream or download podcasts straight from their phones. (Ordinarily, the iPhone can play only the podcasts you’ve downloaded through your computer.) Sokirynsky submitted the program, called Podcaster, for inclusion in Apple’s iPhone App Store—the only way for third-party developers to distribute their programs to iPhone users. Weeks passed; Sokirynsky heard nothing from Apple. Then, on Sept. 11, the company sent him a note. Apple had rejected Podcaster because “it duplicates the functionality of the podcast section of iTunes,” an Apple rep told Sokirynsky.

Apple’s explanation didn’t make any sense. The iPhone App Store carries many programs—for instance, calculators and instant-messenger apps—that mimic desktop software. And anyway, why is “duplicating functionality” so bad—isn’t that the soul of competition? Sokirynsky’s program didn’t seem to violate any written guidelines that Apple had put out for iPhone apps. So why was Apple banning Podcaster?

Nobody knows. In the two months since the App Store’s launch, Apple has rejected several programs for seemingly arbitrary reasons that it won’t disclose. Developers have grumbled about this capriciousness, but until now they’ve had no real alternative—iPhone and iPod Touch owners have already downloaded 100 million apps through the App Store, making Apple the Wal-Mart of mobile software.

Watching Google and Apple carve out space in the mobile business, one can hardly avoid thinking that history is repeating itself. In the 1970s and ’80s, Apple created the first great personal computers. But because Apple closed its platform, it was IBM, Dell, HP, and especially Microsoft that reaped the benefits of Apple’s innovations. The Mac’s operating system ran only on Mac computers; Windows ran on lots of lots of different companies’ hardware. This made non-Apple computers both cheaper than Apple’s machines—competition between hardware manufacturers pushed down prices—and more useful, as third-party developers flocked to write must-have programs for Windows. Apple seems to be following a similar restrictive strategy with the iPhone. Already, some developers have threatened to move to Android; Sokirynsky says he’s building an Android version of Podcaster. Hasn’t Steve Jobs learned anything in the last 30 years?

Well, maybe he has—and maybe he’s betting that these days, “openness” is overrated. For one thing, an open platform is much more technically complex than a closed one. Your Windows computer crashes more often than your Mac computer because—among many other reasons—Windows has to accommodate a wider variety of hardware. Dell’s machines use different hard drives and graphics cards and memory chips than Gateway’s, and they’re both different from Lenovo’s. The Mac OS, meanwhile, has to work on just a small range of Apple’s rigorously tested internal components—which is part of the reason it can run so smoothly. And why is your PC glutted with viruses and spyware? The same openness that makes a platform attractive to legitimate developers makes it a target for illegitimate ones.

Google’s Android OS is “open” in two distinct ways. First, Google has released the software under an open-source license, allowing hardware manufacturers to customize Android for different phones. Second, Android is open to third-party apps; Google and the carriers will make sure that apps do not violate the law or harm people’s phones, but other than that, they promise to impose few restrictions. While this is just what developers like Alex Sokirynsky want to hear, it’s not obvious that this level of openness will be good for users. Will a game that was developed for a phone with a relatively fast processor crash on a phone with a slower processor? What if you buy an app that requires a full keyboard, but you’re running a phone without one—how will Android respond? Engadget pointed out that even before its public launch, Android’s Marketplace is full of programs that don’t adhere to a single design paradigm, making for a “sea of mediocrity.” Is that the danger of running a store without a rigorous approval process?

Let’s remember, too, that if keeping a platform closed was Steve Jobs’ greatest blunder, it was also a part of his greatest success. The iPod is a closed platform: The device runs Apple’s software, it connects to iTunes on your computer, and its music store sells songs that will work only on Apple’s devices. In 2004, Microsoft tried to take on the iPhone with a more open music strategy called PlaysForSure. (OK, relatively more open; PlaysForSure, like songs sold on iTunes, included a copy-protection scheme.) Microsoft set out to certify and license music players made by lots of different manufacturers, an attempt to fight Apple’s music business the same way it had taken on Apple’s computer business. But this time Microsoft failed—the iPod-iTunes link simply worked better than the tangle of PlaysForSure devices. Eventually Microsoft went Apple’s way—it ditched PlaysForSure in favor of the Zune, for which it designs both the hardware and software.

Why did a closed platform hurt the Mac but not the iPod? In his book Inside Steve’s Brain, Wired.com editor Leander Kahney argues that in the early PC market, flexibility was of paramount importance. Businesses were the early adopters of the computer world, and they wanted systems that were cheap and could run all kinds of programs. But those factors aren’t as important in the consumer-electronics market; people like some measure of openness in their devices, but they also don’t mind ceding control in return for security and convenience. The video-game industry proves this. No console is open—Microsoft and Sony approve every game that you can buy on your XBox or PS3. But most gamers don’t mind that, because in return they’re assured that games reach a certain minimum standard—you don’t have to worry that Grand Theft Auto will trash your game system.

None of this is to say there’s anything defensible about Apple’s rejections of iPhone apps. It got rid of I Am Rich, a $1,000 program that did nothing, and Pull My Finger, a fart-joke app, for “limited utility”—which would be understandable if so many iPhone Apps weren’t pretty limited. (How did Apple decide that a program that turns your phone into a flashlight is more useful than a program that turns your phone into a whoopee cushion?) Apple also rejected a comic book app called Murderdrome because its contents were too violent—even though it offers extremely violent movies in the iTunes Store. And it blocked an e-mail client because it competed with the iPhone’s built-in e-mail app, a transparently anti-competitive move.

But there’s no sign, yet, that any of these moves poses any harm to the iPhone’s long-term prospects as an attractive platform for developers. Reports suggest that successful iPhone developers have made hundreds of thousands of dollars—which seems reason enough for them to labor under Apple’s unpredictable policies. Apple seems to be pursuing a strategy of just-open-enough—permissive enough to keep programmers writing code and to keep customers buying software but still locked-down enough to let Apple control the platform’s larger direction.

So far, Apple’s restrictions on apps haven’t turned off iPhone users, either. That may change if someone develops a killer app for Android that can’t be replicated on the iPhone—Skype, for example, or a program to “tether” your phone to your laptop, letting you get online through your cell plan. But if users threaten to quit the iPhone because it lacks certain apps, you can bet that Jobs will find a way to respond. Over the years, he’s shown a willingness to embrace openness when it has suited Apple’s bottom line—the iPod, after all, works on Windows computers. Until that day comes, the iPhone will remain semi-open for business.