What's more insecure, the iPhone or Apple?

It's been interesting to watch the reactions to Apple's crackdown on people who hack their iPhones.

If you've been living in a cave or otherwise off the net, I should explain that Apple's latest software update for the iPhone tends to disable phones that have been hacked to undo the SIM lock (enabling them to make calls on other networks) or to install third party applications. In some cases, Apple has refused to repair the software in these "bricked" phones, forcing the user to buy a new one.

I've read contradictory reports on what level of hacking causes the iPhone to be disabled. Some reports say the update disables the phone only if the SIM lock has been broken. In phones with an intact SIM lock but third party applications, word is that the update "merely" erases the apps without disabling the phone. But the fear among iPhone users is that doing anything unauthorized with the phone, even installing an app, can cause it to be disabled. Apple appears to be feeding this fear deliberately.

This has stopped (at least temporarily) the rapid growth of third party applications that developers and enthusiasts had started creating for the iPhone. Although Apple doesn't endorse or encourage the creation of native apps for the iPhone, developers had quickly found ways to access the modified version of Mac OS X inside the iPhone, and were busily producing a series of interesting and cute add-ons.

I was astounded by the speed at which iPhone applications were appearing. Usually it takes about six months to get developers cranked up on a new device, and that's when things are going well. Just three months after the first shipment of the iPhone, there were already a lot of interesting apps appearing, and David Pogue at the New York Times had even created a video celebrating them (link).

Most technology companies would kill to have that publicity and a bunch of third parties creating new software for their products. Web 2.0 companies are all adding application interfaces so they can get developers, companies like Adobe, Microsoft, and Google are competing aggressively to create APIs for web development, and even Apple invests heavily in encouraging developers to create software for the Mac.

The assault on hacked iPhones has provoked a nasty reaction online, starting among enthusiasts (check out the video here) and now spreading to the mainstream press. The latest example, pointed out to me by Chris Dunphy (an angry iPhone user), is from BusinessWeek (link):

"Wasn't Apple itself the creation of two guys in garage with a knack for making interesting ideas into real things? So why punish the people who try to create something interesting, threatening them with the prospect of an inoperative phone?....The company that styles itself as the technology supplier of choice for creative people with great ideas is insisting that to own its products is to accept a defined orthodoxy where there's only one acceptable way to do things. That doesn't sound like the Apple I know. So I'm not going to buy an iPhone. And until Apple commits to changing this ridiculous policy, I don't think you should either."

I can't remember the last time someone at BusinessWeek actively campaigned against a product of any sort.

Why would Apple expose itself to so much criticism?

The weirdest thing about this whole saga is that it's not at all clear why Apple is putting itself through it. I've been asking myself that a lot, and want to share some thoughts.

The first thing I think we have to do is separate the SIM lock issue from the applications issue. They are two very different business and technical issues, and Apple may have completely different motivations for pursuing them.

Why defend the SIM lock? Many mobile phones, especially in the US, are locked for use on a particular network. All CDMA phones outside of China are like this (because there is no SIM card), and many GSM phones in the US are as well. The excuse for this is usually that the operator paid a subsidy for the phone hardware, and needs to recover the subsidy through service charges. But the operators also achieve this recovery through big cancellation fees if you switch operators before the contract is up, so the industry has not traditionally worked very hard to defend the SIM lock. Unlock codes for many phones are available online, and many operators will reportedly unlock your phone if you call them and say that you're traveling overseas.

Apple is the first phone hardware vendor that I've seen aggressively defend the SIM lock, and I'm not sure why. The most common explanation on the Web is that Apple's getting a revenue share on the monthly billings from iPhone users, so it actually loses a lot of money when any iPhone moves to another network. There is also speculation that if iPhones can be moved into countries where they are not available, Apple will have trouble extracting lots of money from local operators who sign up to carry the phone.

The latter explanation doesn't hold a lot of water for me -- most people want their phone to work in their native language, so an English-language version of the iPhone is not going to destroy the market for a legitimate iPhone in France. Also, iPhones moved onto unauthorized networks lose some of their cool features, such as the visual voicemail function. If Apple were selling iPhones in some countries for $99 and in others for $699, I would see more of a gray market threat, but the price gaps are not nearly that large. Combine the language issue, loss of features, and low opportunity for price arbitrage, and I don't think there is enough motivation for Apple to subject itself to the abuse it's taking.

But the revenue opportunity is a different thing. If Apple got, say, 20% of the mobile billings for an authorized iPhone, that would probably be about $120 a year from an average user -- in pure profit. That's going to be similar to the total margins Apple makes on the actual iPhone, and they get the billings every year. I have no idea if Apple's actually getting 20%, but that sort of number has been rumored for some of the European iPhone deals. Even if Apple's cut is only $10%, the revenue share would be a huge part of Apple's total profit on the iPhone, and something they would be willing to defend vigorously, even if it pisses people off.

Why kill third party applications? This one is harder to understand, because I don't understand what Apple gains from it. Having applications for the iPhone makes it more popular, and also sucks up developer activity that could go to competing products. My first reaction when I heard that Apple wouldn't allow applications on the iPhone was that it was a control issue for Steve Jobs - he watched the base of cool Mac developers get sucked away by Windows, and never wants to be vulnerable to a third party again (link).

There are a lot of commentators online who assume the control freak attitude is driving Apple's behavior on the iPhone. Others speculate that Apple is planning to offer a third party applications store, in which it will take a large revenue cut for third party applications that have been approved by Apple. I have no idea what the cut would be, so it's hard to say how much it's worth to Apple. But I think if it were a big part of their plans, they would have made that store available on the first version of the device. So although I believe they might create such a store (it's an obvious thing to do), I don't think that is the whole explanation. It's hard for me to see them bringing this level of criticism on themselves just to defend that hypothetical store.

Instead, I'm starting to suspect that they have a deeper motivation that they don't want to discuss in public because even acknowledging it could damage iPhone sales. It's better to take criticism from people who think you're evil than to admit that your device has a serious flaw, and I think maybe the security structure of the iPhone is a serious flaw.

When the iPhone was announced, Steve Jobs said it didn't allow third party apps because they could bring down the phone network. I thought that was stupid bluster at the time, because on most smartphones it's very difficult to do anything really nasty to the network. The applications and the phone run on separate processors, and given the limitations of the smartphone operating systems, it's very difficult to do anything really heinous to the network.

But the iPhone has a much more powerful OS in it, a derivative of Unix. The reports posted online by hackers who have played with the innards of the iPhone are very disturbing (link). Here's a great example:

EDGE network access is horribly slow, but it works....I made a few attempts to discover other hosts in the private address space, in hopes of finding other EDGE devices, but instead only found a few scattered routers, switches, and servers.

So the hacker was looking to hack other phones via AT&T's Edge network, and was not able to do so. That's a good thing from the perspective of the average user. But you have to wonder what those "scattered routers, switches, and servers" are. I doubt AT&T deploys switches and servers on its network just for laughs, so who knows how important they are to the functioning of the network, or how secure they are. I'm sure they were not set up with the expectation that hackers would be tickling them from an iPhone.

If you know the technical details of Edge and have any thoughts on this, please post a comment. Maybe I'm overstating the risk here. My personal reaction was that if I worked at an operator and read the quote above, my hair would stand on end (if I still had any).

Here's another interesting quote:

Every process runs as root. MobileSafari, MobileMail, even the Calculator, all run with full root privileges. Any security flaw in any iPhone application can lead to a complete system compromise. A rootkit takes on a whole new meaning when the attacker has access to the camera, microphone, contact list, and phone hardware. Couple this with "always-on" internet access over EDGE and you have a perfect spying device.

Well, that's pretty straightforward. There are already third party applications that turn a smart phone into a spying device, but you need physical access to that particular device in order to install them. The difference with the iPhone, according to this report, is that once you find a security hole you could install that sort of spyware remotely, via the wireless connection.

That led to a Computerworld article which says basically that viruses and other malware could spread from one iPhone directly to another without the user ever being aware of it (link). I'm not too alarmed by that just yet, because there isn't a critical mass of iPhones in any one geographic location to infect each other. But it could be interesting the next time there's a big gathering of iPhone users. Macworld, anyone?

To me the more troubling part of the report was the root privileges thing. I'm not a Unix expert, so I talked to someone who is. He confirmed that applications with root privileges in Unix can do just about anything. Unix is designed to empower programmers, and the assumption is that someone with root access knows what they are doing and can be trusted. (You can read some similar commentary in a eWeek column here).

There are ways to prevent third party applications from having root access, but the disturbing possibility (and I'm speculating here) is that Apple may have stripped out those protections in order to reduce the memory requirements of the iPhone and make it run faster. If that's the case, my friend said, it may be a pretty involved project for Apple to add those protections back in. Not at all impossible, but requiring a lot of work and time.

Through my years in the industry, I've done a lot of research on technology users. One of the things I've learned is that security problems are a great way to scare people away from a new technology device. If it even sounds insecure, a lot of people will stay away from it. Based on what I'm seeing online, there is a lot of evidence that the iPhone as currently structured is a genuinely insecure device once any uncontrolled third party applications get onto it. What's more, keeping third party apps off your own iPhone does not necessarily protect you, because malicious software could propagate from device to device.

If I were working at Apple, and this were the situation, what would I do? Well, first I would not want to acknowledge the vulnerability, because that itself would scare away customers. Second, I would do everything in my power to shut down all third party native application development. Squash it, kill it completely. And I'd be willing to take a lot of criticism for doing so because the alternative, acknowledging the security problem, would produce even more bad PR.

Let me be very clear here: I'm not saying that I know this is what's going on at Apple; I don't. And I'm not trying to start any nasty rumors (they are already out there). I should also point out that some reports on iPhone security have been a lot less alarmist (for example, here is Symantec's take from early July). But that was before the latest reports surfaced.

I think we need to ask whether Apple botched the security of the iPhone in the belief that people wouldn't try to add apps to it. They could easily have made that assumption; there have been comparatively few efforts to add apps to the iPod, after all. But the publicity for the iPhone, and Apple's bragging that OS X was in it, made it an irresistible target for hacking.

If Apple really does have a security problem in the iPhone, I don't think they will be able to keep it quiet. Experience shows that the best approach in this sort of situation is to come clean about the problem, take your lumps, and fix it as soon as you can. That way you at least retain your reputation for honesty. If the iPhone really is vulnerable, Apple risks ending up with the worst of all possible worlds -- it'll damage its reputation for honesty, piss off a lot of technophiles, and people will still hear that the iPhone is insecure.

It will be interesting to see how Apple handles this issue in the weeks to come.


Thanks to John Hering at Flexilis for pointing me to the Computerworld story.


menneke said...

Great article!

SamiU said...

Good summary on recent stuff.

The "third party apps could bring down the phone network" argument is not really valid. You can connect laptops to the EDGE (and 3G) network with datacards and the pipe from you laptop to the network will be exactly the same as from your iPhone to the network; i.e., the same routers and other gear should be visible also from your laptop and hopefully ATT and other operators have those protected properly. Anyhow, the network gear visible from user devices are actually not part of the cellular network, but part to the operators internal IP backbone. Thus, attacks to those devices should not have impact to the cellular network - attack on those could impact the Internet connectivity, but should not have impact to voice and SMS or other cellular functionality.

You can have shell and all the other tools running in S60 devices and this has not been considered as a security threat to cellular networks...

However, if the iPhone's whole security model is flawed and users can modify the way how the phones cellular stack behaves, this is more serious and may indeed cause some issues for operators, but very unlikely to "bring down the network"

Anonymous said...

Thoughtful commentary, as always Michael. Thanks. While I suspect security is one of the issues Apple is struggling with, they really haven't done that much to block native iPhone app development. I think the primary reason they're discouraging it for now is that they just do not feel they have an SDK that they are ready to support for production quality applications. But they do have one that's good enough for Yahoo and Google coders to make nice use of.

Establishing a new platform, borrowing Leopard resources, entering a new competitive space, that's a lot to juggle. Their primary concern is delivering a usable product up to people's expectation of a phone & iPod. I expect that a year down the road they will have an SDK. Isn't that about how long it took the original Palm OS? I don't recall an SDK from them right out the gate either...

Dave Haupert said...

Good commentary. A few thoughts to add:

1. I have long believed the reason for no third party apps is that they don't want Microsoft and Real to create DRM protected music players for the iPhone (or iPod Touch for that matter). If they did open up for 3rd party apps, they'd now have Napster and Rhapsody, etc likely releasing the ability to buy/rent/download/etc songs from their networks, and breaking the locks of exlusivity that have frustrated everyone but Apple for the last 7+ years. Just my hunch.

2. If they went with a certified apps route, they could control and/or mitigate the chance that the 'entire network' could go down. So the question is why they aren't going this route if that's truly what they believe. Obviously they don't really believe it!

3. This whole root access of all built in apps has already caused a full overtaking of the 1.1.1 system, from what I've read online. They found a vulnerability in Safari in rendering TIFFs and used that to overtake the system, and thus all third party apps and unlock software works again with the latest software. This was as of yesterday, probably as you were writing this article!

Anonymous said...

Another semi-related comment: The security (or lack thereof) of the iPhone is the perfect string for RIM to pull on, from a competitive point of view, since BlackBerry security is one of RIM's distinctive competences. The question is, will they? RIM has been notoriously passive at going after their competitors. It will be interesting to see what will happen, but in the light of these new reports, I'd say RIM stock is probably a safe bet :).

Stuart said...

Regardless of the reason for Apple's position on third-party applications, I think there's a bigger point here to be made about security models for mobile computing.

I recently read a comment on Slashdot which pointed out that today's user-based desktop security (Windows, Mac, Linux) is based on the model of an all-powerful, benevolent system administrator (root) managing a set of users sharing computing resources. The main purpose of user-based security is to provide privacy and controlled sharing to these users, and to prevent these users from bypassing the security decisions made by the adminstrator.

Personal computing made this model outdated, and mobile computing has made it completely obsolete. Today, we have computing devices where the administrator and the "users" are all a single person. In this situation, what good does it do to have a security model that tries to protect the user from himself?

Computing -- especially mobile computing -- needs a new security model based on units of code (applications, libraries, plugins) rather than users. We need a system where it is trivially easy to distinguish one application from another, newly downloaded software from pre-installed software, and software created by different vendors. And then based on these distinctions, to allow different levels of access to computing resources.

I believe that Symbian and Windows Mobile do some of this using digital signatures, but the concept really needs to be refined to handle things like self-signed code and untrusted plugins in trusted applications. And to be truly secure it really should be part of the kernel rather than layered on top of it.

Apple may have realized that user-based security was serving no purpose on the iPhone other than to slow things down -- so they got rid of it. Now we need to see them replace it with something that actually allows third-party software to run safely on their device.

Patrick said...

Very interesting read, Michael, as usual!

I'm not a unix specialist, although I've "played" a bit with Linux some time ago, and from what I was able to understand, one needs to "jailbreak" an iPhone to alter it in any significant way.

In other words, you might get read access to everything by hacking your way onto the system, but for now, at least, getting write access is a bit more complicated if you want to install third party applications - or a virus/worm.

Take a third party like "Installer", the application which, once installed on your iPhone, lets you install dozens of other third party apps from a no-brainer GUI.

Well, if your iPhone is not in the aforementioned "jailbroken" state, i.e. with write enabled, then you can't install a single application... (I'm not sure how many people leave their beloved iPhone in that insecure - but very handy - state when leaving home in the morning).

And from what you describe, due to the lack of security scheme for third party applications, they certainly all run with root privileges as well (the lighttpd web server you can install on your iPhone, for instance, is listed with a "root" user and the "wheel" group), so if the iPhone isn't jailbroken a virus can probably do only a limited amount of damage. Not more (and probably less) than what could be achieved by any computer or "open" smartphone connected to the internet via a GSM/GPRS/EDGE network.

Not that it wouldn't protect people from having one's iPhone "taken over" by surfing a website containing a cleverly designed TIFF image, but rebooting your iPhone and emptying the cache would erase the offending application, I'd say.

This leaves room only for malicious pictures stored in the (always writable) Media folder, but since you can't yet save an attached picture to your Media folder, it means you'd have to put it in that folder using iTunes. Not the easiest way to spread a virus, I think.

Also, I don't think there's a special reason why a gathering of iPhones would make them more vulnerable. It certainly seems to be easier to attack an iPhone connected to a malicious website than trying to propagate an evil application directly from device to device, as I'm not aware right now of any way to make two iPhone communicate with each other, except via the GSM (using SMS, but images or objects cannot currently be sent via SMS) or internet (using Wifi or EDGE, but then the iPhone would be connected to servers, not directly to your neighbor's iPhone) networks.

There's just no IR capabilities and the bluetooth stack doesn't handle object exchange from what I could figure...

Maybe I'm missing something obvious, though, those are just a couple of thoughts provoqued by your inspiring article, not the result of a rigorous analysis! :-)

Jason Devitt said...


Great post, and you may be right about the security model. Nokia and Microsoft have invested a lot more money in security than Apple has and they still haven't got it right. (Patrick: Michael is speculating about the risk to Apple of permitting third-party apps today; given that they do not, the actual risks are minimal.)

I think you are underestimating the revenue that Apple and AT&T stand to lose if the platform is truly open. Dave brings up Real and Napster; recall how Apple fought back when Real tried to hack the iPod. The market for premium content and apps on mobile phones is around $20 billion globally, far greater than the market for music online. Without some restrictions on how you transfer content to a phone, that market would collapse.

It's not just content though. Why hasn't Apple ported iChat to the iPhone? I wrote about this the week after the launch: $5,000 per megabyte.

Anonymous said...

I think there could be a lot we do not know about unlimted data usage in general from the telco prespective. I would not be all that suprised if apple's share of revenue decreases if too much traffic starts showing up on AT&T's network as a result of heavy data usage.

I am starting to believe that a good part of the limitations that operators put in place for their data networks are not about bussiness model as much as preventing network overloading. I am thinking this since I have taken some time to study how these network operate and their capacities.

Michael Mace said...

Thanks for all the comments, folks. This is exactly the sort of discussion I was hoping for.

Observations on a couple of the comments...

SamiU wrote:

>>You can connect laptops to the EDGE (and 3G) network with datacards and the pipe from you laptop to the network will be exactly the same as from your iPhone to the network; i.e., the same routers and other gear should be visible also from your laptop and hopefully ATT and other operators have those protected properly.

OK, good explanation, and I feel better about that one now. Thanks.

Anonymous wrote:

>>I expect that a year down the road they will have an SDK. Isn't that about how long it took the original Palm OS?

No comparison at all. Palm was a tiny startup; Apple is one of the most successful tech companies in the world.

Anonymous wrote:

>>The security (or lack thereof) of the iPhone is the perfect string for RIM to pull on, from a competitive point of view, since BlackBerry security is one of RIM's distinctive competences.

You're right technically, but I think the Blackberry sells to a very different audience than the iPhone. A Blackberry is a lousy media player and an iPhone is not an Outlook client at all.

Stuart wrote:

>>I think there's a bigger point here to be made about security models for mobile computing.

Very nicely argued, Stuart. Great comment.

Patrick wrote:

>>I don't think there's a special reason why a gathering of iPhones would make them more vulnerable.

It's the same reason why flu germs like airplanes -- when you have a lot of susceptible carriers crammed together it is easier for the virus to spread quickly.

>>It certainly seems to be easier to attack an iPhone connected to a malicious website than trying to propagate an evil application directly from device to device, as I'm not aware right now of any way to make two iPhone communicate with each other, except via the GSM (using SMS, but images or objects cannot currently be sent via SMS) or internet (using Wifi or EDGE, but then the iPhone would be connected to servers, not directly to your neighbor's iPhone) networks.

I think the scenarios being tossed around online center on Bluetooth, but I'm reaching the limits of my knowledge.

Anonymous wrote:

>>I am starting to believe that a good part of the limitations that operators put in place for their data networks are not about bussiness model as much as preventing network overloading.

I've had more than one developer tell me that they think the operators are schizo on data applications -- they want users to install them but then never use them.

Oh Blah Dee Blah Dah said...

RE: "Why kill third party applications?"

You gave very good reasons why Apple SHOULD kill native applications.

The iPhone is on a NETWORK and a "root user" has the capability to do damage. We don't need to wait for critical mass of users to exist. Caution should be taken from the beginning, and I strongly agree with Apple's approach to going with web applications.

Earlier in the U.S. radium water, yes RADIUM water, was sold in stores as a beneficial drink. After the danger of this product was discovered it had to be removed from shelves and banned.

The above example is not exactly comparable, however, we don't need a major network disaster, with the possible loss of life, to realize that hackers have done damage WORLDWIDE using computers, and that the Apple iPhone IS a (handheld) computer that could be hacked.

alex said...

I also believe in "not yet ready for an SDK" as the main technological reason. Apple's experience is in PC-centric APIs and it will take them a while to build up mobile competence before they can offer stabile APIs.
Just look at Symbian/S60 APIs, which loose backwards compatibility at every major release. Symbian companies have mobile competence but seem to still lack design competence for 3rd party developer APIs.

As the main business reason, I would venture the deals with the operators. Apple got Wi-Fi into AT&T (quite remarkable) and in return AT&T does not have to fear rogue VoIP, IM, and other "operator by-pass" applications either on the cellular network or Wi-Fi.

At least equally important for AT&T is probably the more predictable and lower data usage if application deployment is tightly controlled - balancing low cellular capacity with flat-rate data plans.
Note also that iPhone doesn't support Flash, Java, and other platforms which are often used for bandwidth-hungry appliations, whereas YouTube on iPhone is implemented bandwidth efficiently.
I could well imagine that Apple signed up with AT&T to "x years of no native APIs".

This plays out well for Apple also internationally. They need bigger operators for their initial deals to create an international footprint, and the bigger operators are usually the more restrictive towards "harmful" applications.

This topic is also interesting for your Apple vs. Nokia comparison.
While Nokia is trying to move away from the operators as much as it can afford to (building direct device sales channel, OVI service platform, etc.), Apple has chosen to work with the operators as much as it can afford.


Anonymous said...

we will get much higher usage of mobile services long term if everything is browser based; not apps loaded onto the phones. apple has the right idea here. people do not think of their phones as PC's. they do not want to install applications on them. they want to use them to log onto web pages and use interesting services on a one off or longer basis. when they change to a new phone they should not have to port or reload application; it is better if they can just log onto a site.

i know that there are technological limitations that make phone hosted apps attractive. to get widespread adoptation these limitations need to be overcome. we need HSDPA/EVDO rev. A everywhere and high capacity pipes on the backhaul networks.

this not only applies to phones. i find many people will not download anything on their PCs over the web; these same people will try anything and everything that they can open in a browser window without that windows installation wizard poping up. these people will not want that the menus on their phone have suddenlly been altered by a downloaded app; they will however be fine with bookmarking a website for a later return visit. increasing amounts of spyware/adware only put more people in this group of non-downloaders/

Michael G said...

Anonymous said "Apple has the right idea here. People do not want to install applications on their phones... I know that there are technological limitations..."

To some extent, I agree. I have very little installed on my Treo, less than I had on my previous Palms. I get my email, train schedules, etc. from the web browser. In the past, I used native apps for these.

For other apps, like calendar/contacts/memos, the wireless connection isn't fast enough or ubiquitous enough, yet.

Kevin said...

Well, Apple has announced an SDK because it is now a sure thing technologically speaking, and they have a release date. Apple doesn't announce anything in advance unless they have a release date, and then they only pre-announce (Intel switch, AppleTV preview, Leopard preview, iPhone preview, SDK) if it has some benefit, i.e., preparing developers, signing up content partners, preparing consumers by not having them lock in to new contracts.

As for your musings, I just don't know why people (and you) don't listen more carefully. Jobs said in Jan 2007 (John Markoff article in NYTimes) what Apple was thinking and working on. He amplified on it at All Things D (see transcript or highlights at website), saying be patient, we're working on it. In interviews after the June launch, he repeats it again; be patient - obviously, Apple thinks security and protection are crucial to the user experience of the mainstream user (their target audience). Did you know nearly 25% of iPhone owners are between 50-60 years old? And now in his letter, he gives the same consistent explanation. It's clear that Apple saw a huge security issue, and had no intention of becoming the Windows/IE nightmare for the mobile world. (I'll give you credit for musing a bit on what and how big this security issue could be.)

So now that OS X Leopard is done, and includes the necessary underpinnings for an iPhone SDK, they announce. As Apple thought it the needed features could miss Leopard's deadline, they could not pre-announce an SDK, because it might not have come for a long time. All they could say is we're working on it, which is what they said.

It's clear iPhone 1.1.1 made significant changes, and likely more big changes, related to the finalization of Leopard, are coming before SDK release. If nothing else, this is why Apple doesn't want anyone writing apps today; Apple is still busy changing the foundations.

Kevin said...

Regarding unlocking and revenue share, I think Apple's bigger strategy is to help AT&T switch from voice+data plans to all-data (including VoIP) plans, which could be priced by speed and data limit tiers when we get to 3G and beyond. So today, it's $20 for "unlimited" data on EDGE; in the future (likely when 3G+ arrives in US), it could be $60, $80, $100 for various tiers of data limits and speeds.

Thus, in exchange for revenue share, Apple has promised to bring more customers to AT&T, and to convince those customers to pay for data; most of whom have never paid for data. The handwriting is already on the wall that paying for cell voice will go the way of landline voice.

Apple thinks iPhone has the best interface for using data, and so will be adding data-driven apps (including VoIP) to get more people to sign up for higher cost data plans. And developing an SDK to bring even more data-driven apps (but without sacrificing security). And this is why Apple thought Web apps was mucho good too.

Apple needs AT&T to be successful so unlocking must be stopped. AT&T's success will lead other carriers to also shift and stop being "orifices" and "walled gardens". And that also benefits Apple's dominant iTunes store.

I don't think the revenue share is that important long-term to Apple; it's leverage for now, possibly to be traded away when 3G and the all-data plans arrive.

Michael Mace said...

Good comments, folks. I especially like the discussion of hosted vs. native apps on phones. You summarized the situation well.

I wanted to comment on a couple of Kevin's points...

>>I just don't know why people (and you) don't listen more carefully

I guess listening is a subjective thing, but I thought I was listening pretty carefully, and what I heard was Apple posturing that Ajax applications were the right (and only) way to create apps for the iPhone. From my perspective, it appears that Apple changed its story after third party developers forced the issue. But perhaps you have more insight into Apple's internal thinking than I do.

>>I think Apple's bigger strategy is to help AT&T switch from voice+data plans to all-data (including VoIP) plans, which could be priced by speed and data limit tiers when we get to 3G and beyond. So today, it's $20 for "unlimited" data on EDGE; in the future (likely when 3G+ arrives in US), it could be $60, $80, $100 for various tiers of data limits and speeds.

Could be. I think the price is actually $60 if you're using a PC, which is more like the traffic volume I'd expect from an iPhone (details here.)

If the iPhone really takes off and starts driving a lot of multimedia downloading, I wonder if AT&T's network can stand up to the traffic. A lot of people have said that even 3G networks don't have the capacity for a lot of people to use video...

Anonymous said...

apple is going credit card purchase only on the iPhone; no cash. that has got to be one of the weirdest things I have ever heard. i believe it has no chance of working in europe where credit cards are not popular and cash is still the preffered way to recieve payments by everyone. so all the grey market iPhones will be coming from over there or the resellers will just use dissposible prepay credit card(unless apple refuses to accept thoose) what I really do not understand is that apple does not seem to want to sell these additional iphones(i say additional because I do not believe that many people who buy iphones and do not activate on AT&T would use AT&T as the alternative; they would choose a different phone. the few people I know who have unlocked iPhones have a few different smart phones and switch them out. its really a hobby for them. very different people than the locked iPhone crowd)


Anonymous said...

after vodaphone in germany brought a lawsuit against t-mobile the courts ordered t-mobile to sell unlocked iphones. you can now walk into a t-mobile shop in germany and buy a brand new unlocked iphone for 999 euros(versus 399 locked) thats about $1500.

anyone want to speculate on the gray market value?

Anonymous said...

iphone from t-mobile are unlocked now with or without contract but there is a 600 euro price difference. i am not sure but it may be cheaper to pay a the contract for 2 years even if you do not use it.

apparently it is now illegal to sell locked phones in germany. but it took the iphone and the exclusive deal t-mobile had to take it to the courts.

kathy green said...

Apple has a difficult road supporting 3rd party apps. Hopefully this SDK model they are introducing will keep it cool. Some info here iPhone