don't use your iPhone the way you want to use it...

Yeah, it's all a bit silly. It's a shame that we just justify any invasion of privacy, and inconvenience, by dreaming up a wacky nightmare scenario that it would protect us from. Is there a risk that some malicious app might worm its way onto the iPhone? Yeah, sure. But there's risk in everything we do. There's also a risk that I might get into a car accident driving home from work today, but that doesn't mean I want Apple remotely disabling my car to prevent that. I've got an iPhone, and I love it, but I certainly won't pretend that I love everything Apple does. And this would be one of those things I'm not crazy about.

Disclaimer: I haven't read the WSJ article, mostly because I apparently have to pay $89 to do so. They may present facts that makes what I have to say below moot.

Much hay has been made about Jonathan Zdziarski's discovery of what appears to be a blacklist file in the depths of the iPhone OS. It's important, I think, to get the technical details right, here, because equating what he found with Apple being able to "nuke applications" is somewhat hyperbolic.

Zdziarski found URL buried inside the Core Location libraries that points to data file on Apple's website that looks like it's structured to hold a blacklist of applications. But: There were no applications in the list. There file was found as part of Core Location (the iPhone OS libraries that handle the location-awareness features of the phone -- the GPS and friends), not as part of the general OS. And (and this is important) nobody knows what the semantics of the list are. That is, nobody knows what will happen if an application is on the list.

It *may* be that the application will uninstall itself. But that's not *necessarily* so, however. The effects of the blacklist may just be to disable Core Location services for that application without otherwise impeding it. Or it may force the application to quit until an update is run from the App Store.

In any case, I'm of two minds about the amount of control Apple should exert over this kind of thing. The coder in me wants the iPhone to be a totally open platform -- every service and feature of the phone fully exposed, with applications unregulated by anyone. But there are problems with that (software quality aside). Phones aren't like computers; they have access to lots of sensitive information (who you call, where you are, what you're saying, what the camera is pointing at, etc). Totally unrestricted applications could easily be used for all kinds of nefarious purposes.

So, Apple has put several levels of restrictions in place. First of all, the scope of applications that can be built for the iPhone is limited in a number of important ways : applications can't see each other's data, or run in the background. Access to some services is very limited (the camera, for example), and access to others is subject to overrides outside of the application's controls (Core Location, for one).

Apple's legal terms with its developers also restrict the kinds of applications they can write -- for example, you can't write an iPhone application designed to operate your automobile.

Second, there's an approval process for applications -- you send your application to Apple, they decide if it's OK to go on the store or not. Theoretically, obvious malware could be caught like this. It doesn't look like Apple are doing a great job at this; the approval times for many applications -- and updates to them -- is very long, and they aren't enforcing any kind of concrete quality bar. And, as usual will Apple, the whole process is totally opaque; there's no specific policy on what they will or won't allow in the Store, and nothing to say why an app was removed from the store (even, apparently, to the folks who wrote the application!)

Third, there are (apparently) some kinds of after-sale interventions Apple can do to disable, restrict, or remove software that's already been installed on an iPhone. Exactly what the scope of those interventions are is (as far as I know) unknown. I can certainly envisage malware that you'd want to be able to remotely because it's impacting other users -- a game that runs network denial-of-service attacks in the background, locking out traffic from local wifi access points, for example.

(another possible scenario is where an application is violating intellectual property rights -- should Apple be able to remotely wipe the application, or just stop selling it?)

There's also a risk that I might get into a car accident driving home from work today, but that doesn't mean I want Apple remotely disabling my car to prevent that.

That's not what's happening -- it's more like Apple being able to turn off your spinning hubcaps. Your car is still capable of doing all of the car-like stuff.

(second disclaimer: I work as an iPhone developer (but not for Apple), and probably have a vague pro-Apple bias).

Sorry about the monolithic text in my previous post -- I didn't know I needed to manually insert paragraph tags.

I inserted paragraph breaks for you. I probably should make a note of the need to do that for commenters...

I can certainly envisage malware that you'd want to be able to remotely because it's impacting other users

Well, that's the thing. It's always easy to envisage *good* things that can be done with great power. But it's also always easy to envisage *bad* things that can be done with it.