[CoffeeDregs]

Siri is a nice technology (though, after having seen 3 attempted uses, I have yet to see it work), but the sell-hardware-backed-by-nice-software/services is not some revolutionary business model; further, often, it's not one that is sustainable for long. The problem is that as the non-hardware costs start to become material, the pressure mounts to show that they are actually related to hardware sales; now they are distributed across hardware sales as a per-unit transfer cost and become a cost-center; now investors and organizational politics begin to pay attention and the model of buy-random-shit-and-claim-it-is-accretive begins to sour.

This is why it is so hard to do hardware and software together. There are few examples of long term success in hardware/software and Apple is certainly not one of them (except in the '00s). IBM is one. But, if one area of the business takes off, the pressure to drop the other is immense (see Thinkpad and HP's current flail-ures). As XCode users will recall, there are also issues with adding significant value to past purchasers: if you paid $100 for a product that had $20 of awesome value added later by an upgrade, the company who took your $100 of cash should only be able to recognize $80 of revenue now.

[rflrob]

> The problem is that as the non-hardware costs start to become material, the pressure mounts to show that they are actually related to hardware sales

I think that implicit in Gruber's argument is the idea that just buying Apple's devices isn't radically more expensive than the competitors. I've heard this attributed to Tim Cook, that lots of Apple devices are now roughly price-equivalent with other brands, but with so many other features.

[gujk]

For a company with no plans to float new debt or investment, and a CEO who plans to stay in his job for a decade, short term accounting games have no effect on company performance.