Data specific to an Apple product was stolen, but this isn't at all similar in any way shape or form to the BP situation. That's like saying VISA was breached when Monoprice's servers were siphoned for transaction/billing details.

Further, the data wasn't obtained by compromising an iPad as the story initially leads you to believe as well. That's just Gawker spicing things up.

It's shameful on ATT and yes, Apple can't help but be involved here. But they also have no choice in the US for other partners since all the other networks have their heads up their asses with out-dated, ill-conceived, non-standard and non-compatible 2G and 3G networks. The story can be reported without a libelous slant and misdirection.

Yeah, and to be clear, I'm not disputing that Gawker took some cheap shots, but you went too far in trying to distance the two entities in your post. Apple and AT&T are partners in the US, and, regardless of how and why that relationship came to be, they share responsibility for these issues.

We still disagree on the term "responsibility" - in my view (at least so far as security goes), Apple are no more responsible than the people whose email addresses were obtained. In fact, even less so. Apple didn't give ATT the data, their customers did. It would be no different if data was stolen from Best Buy or Walmart - Apple does business with those guys too. Apple's association ends at their decision to do business with the company responsible - and as mentioned, they didn't really have much of a choice in that regard).

I'm glad to see another blog call out Gawker on their bullshit: TUAW

They didn't have a choice because AT&T's network is more standard? How does that follow?

ATT is the only choice in the US when making a mobile telecom product that's also supposed to cater to the rest of the world. GSM/UTMS, using common frequency bands. Three UMTS bands and four GSM being enough to cover pretty much everywhere, which every iPhone until version 4 was equipped with (minus UMTS for version 1).

Right, and my point is Apple had a choice -- they could have done GSM abroad and CDMA in the US. It's not like other manufacturers don't deal with it. Saying they had no choice just because they chose some arbitrary limitations on that choice doesn't wash. They may have made a financially prudent choice, but they had other carriers to choose from.

No one said it had anything to do with Apple. Defensive much?

Gawker did. They've been trying to defame the company for weeks now - it's so ridiculous it's now being covered by other blogs.

Tony, I don't know of any other single phone off hand that offers both GSM and CDMA. So you're suggesting that Apple could have made two distinct and incompatible (so far as cellular networking goes) products. The US product further unable to be taken out of the country because it would fail to work anywhere else. Yeah, that's definitely a choice. Up there with picking between flying to Europe or rowing a boat while dragging an anchor.

Could have done, but that's then making two separate products which takes 2x the effort... speaking personally, I spent enough time in china for the first phone

Blackberry make the 8830 that can do GSM and CDMA.


The choice is to make it like the Blackberry 8830 World Edition?

There are products that do GSM and CDMA. Look at the verizon world phones (I think they sell a couple of blackberries that fit this bill). The consequence is generally a bigger/more expensive phone as you're paying for, and having to fit, circuitry that you don't need on your "home" network, whatever flavour that might be.

Why is this so radical, exactly? There is a CDMA Palm Pre for Sprint and Verizon, and a GSM Palm Pre for AT&T. I don't follow the mobile product landscape too closely, but I'm sure there are others. The phones don't have to interoperate with each other, just the networks they're on. It's simple, if you want access to more subscribers, you make a CDMA variant of your phone. Apple decided to do the exclusive thing for 5 years or whatever. Nobody forced them to.


There are plenty of people who choose a phone that doesn't work overseas, myself included. I'd rather deal with using a prepaid phone overseas than to be a slave to AT&T.

It's not like you have to rewrite the OS and apps for the CDMA variant, or change a vast majority of the components inside. Just a different radio and some changes to the hardware design, right?

So, not 2x the effort.. maybe 1.05x the effort, or less, probably?

Oh, I just take my GSM phone and put a prepaid SIM in it. The inability to do that with nearly all iPhones irks me to no end and you'd likely read in another of my posts.

Having two products would be a nightmare. Apple is as profitable as it is because it knows how to mind its costs (as well as sell a lot of product). Having two supply chains is a lot of work. Definitely 2x the work for someone that isn't involved with the OS. Hugo is on the hardware side right?

Putting CDMA hardware into all iPhones would have been expensive. Not to mention detrimental to the design which is already using all the space afforded by its footprint. Doesn't seem logical to me to toss in support for technologies that are only used in the US and that are just going to disappear.

Anyway, none of this really has anything to do with the issue that ATT had some poor security on some of the data on its web site.

Maybe it's more work for the team that does the hardware design, though a majority of the hardware components would be shared. What you get for that extra work is access to Verizon and Sprint's customers. There are a lot of them.

Obviously, the lucrative exclusive contract is the reason they didn't go that way. Were it not for that contract, there would be a CDMA iPhone, and it would be on Sprint and Verizon. Making your phone available to more subscribers is going to be worth the extra cost and duplication of supply chains and yadda yadda.



No, but you made a false statement about how Apple had no choice other than AT&T. They had plenty of choices, but they picked the lucrative AT&T exclusive deal.

Ok... They had no other "sensible" choices at the time.

It was only sensible to go with AT&T because AT&T waved a giant stack of cash and it smelled good to Apple and its shareholders. The other choices were sensible, just not as lucrative.

I don't begrudge them for pursuing the best financial deal possible, but you can't then say they had no other choices because of technologies. It was a financial decision, not a technological decision, and if Verizon were coughing up $400 per phone (or whatever it is) the iPhone would be CDMA-only in the US and GSM everywhere else.

I had no idea Tony had such inside connections into Apple's business and contract decisions to know more about the situation then an Apple hardware engineer

Errrr, no.

Well, not from a HW point of view at least; you're talking about a totally different RF hardware team (nothing is common, pretty much) then different baseband software stack, maybe different software interface to the applications processor, a new RF test team (different specs, different standards), a new approvals team to deal with the approvals, then every test you run on the hardware from reliability to EMC to ESD has to be run on two different designs.

If you look at eg, how motorola is structured, they have totally different teams working on CDMA and GSM variants of their phones. It's not like one team cranks out both; a platform team will make the AP section then two separate RF platform teams will contribute the RF platforms, then two separate project teams will glue bits together and deal with all the integration issues.

Can it be done? Obviously it can be done. Motorola and HTC do it all the time. Ever noticed how Apple tends to concentrate their effort on one thing at a time and do it really well instead of trying to please everyone?

Look at it in that context and you'll see why if you're picking one thing to make, you don't make a CDMA phone. It's needlessly limiting. Make a WCDMA/GSM phone and you can sell it everywhere including the USA.

Apple signed on with Cingular prior to ATT's acquisition. I doubt very much there was a big stack of cash waved up front. What there may have been was a really good subsidy and bonus model and of course, as mentioned, the huge benefits of a GSM network.

Tony, you don't think Apple could have come out with an aPhone that ran the same software as the iPhone but didn't violate the agreement with Cingular/ATT? I don't think the exclusivity contract would have stood in the way if Apple saw a large enough benefit to having previously supported Verizon.

Verizon and the other US carriers need to roll out new networks. Bell has done just this in Canada, bringing along Telus for the ride. Once the exclusivity is over, I'm confident we'll still see Apple selling a single handset. This seems to be working quite well for Apple so far.

I enjoy the way you transitioned the discussion from an AT&T security breach, to defending Apple to all the Gawker editors that read empegBBS, to why Apple's decision to choose AT&T as their sole carrier was the only possible choice.

<golf clap>

I'm only trying to reply to many of the points being made in the obviously evolving thread. Should I just ignore what other people are writing and simply repeat what I've said in a previous post to stay on topic?

This is an iPad thread and Tony's link about the iPad owner email addresses was totally relevant, however the quote and the originating story (at Gawker) were a giant pile of slanted crap. I only wanted to point out that bit of trivia and that this was squarely an ATT web site security issue.

Actually, this is an AT&T data plan/maybe the Evo looks nice/GSM and world phone thread. The iPad/ebook/flash/iAds that don't run on the iPad thread is titled "iPad". All of the above seem to be nitpick Bruno threads

*ponders just turning off thread titles*

Yeah, sorry... ATT thread. Making Tony's link and the "real" story mostly on-topic (since it's ATT's ineptitude and all).

Ultimately, I defer to your expertise on the issue, but your statement that it's 2x as much effort wasn't limited to the hardware aspects. I get what you're saying about having separate test teams, some duplication at the software level, and people to deal with regulatory crap, and I probably lowballed my 5% higher number, but I just can't see it being 100% higher. Maybe there's not as much shared as I thought, but is it true that there is zero shared?

Based on his comment of "totally different RF hardware team (nothing is common, pretty much)" I'd assume not much can be shared, though I am curious too. Beyond just designing the phone, you also have the overhead of manufacturing two separate mainboards (based on photos showing the radio/SIM bits on the same PCB as the flash/processor), sourcing the different parts, additional logistics or shipping more SKUs, and so on.

I'd actually be interested in seeing a teardown of one of the Chinese iPhones (the ones lacking WiFi) to see how Apple currently handles that. With those phones it's just a matter of taking something out.

The iPad would likely be easier to go CDMA first, with the modem being a separate PCB that plugs into the mainboard. And with no voice, simultaneous voice/data issues go away.

Easier perhaps, but there's even less reason to offer a CDMA alternative for that platform for a number of reasons. The first one is that the existing service plan available for it isn't tied to your (any) existing cell plan. Presumably having a CDMA board would mean other carriers would/could offer something similar, making the only advantage one of possible price competition for that data plan.

The iPad is more of a computer than pocketable communications device. Chances are you'll have (and want) WiFi where you'll be using it most. A lot of people are going to skip the 3G add-on. I'm not sure creating two different add-ons is going to bring double the upgrades.

We don't deal with hardware designs nearly as complex as the iPhone, but for us, even seemingly minor changes can introduce a massive workload in hardware design.

You know, when you're fitting in large pin count BGA devices into as tighter space as possible you make use of every single last space on the board as possible, even so much as adding a new resistor or changing swapping the functionality of a pin on the device can introduce days to weeks of work for the layout people.

In Hugos world things are even more amplified by having so many different requirements (certification, testing, manufacture etc).

Given whats already been said as you point out they just won't fit the components that are required for WiFI, presumably the software doesn't need a special build as it'll know from its model number whether or not to initialise and provide options for WiFI to the user.

We have our own pick and place machine at work now, and I modified our cad software to support build attributes, therefore every component on the board can be part of a specific build (and have a specific value for that build), it outputs direct to the pick and place machine. It's literally a case of "Select which options you want on this board and then upload to the machine", press start and the machine goes off placing all the components you've selected.

We try to look as far in advance as possible, what our customers are likely to want (we're talking oems here) and we build in that functionality, it's quite nice when they say "actually, could you make it do x? sure, we made provisions for that in the design". It always happens and most of the time it's a doddle to accomodate what they want.

I am obviously coming at it from the HW point of view; much SW would be shared, but software already developed just kinda stays developed and can get shuffled onto new platforms as required - there's very little effort required there.

On the HW front it's multiplication of effort. Every design is different, and it's not as if there is an infinite pool of the right engineers already employed either so it's much harder than saying "this is the direction we're going to go in".