To have ARM based Desktops and Notebooks is a sweet deal, I truly like this architecture, it is solid and powerful. The thing that scares me most is a computer based on IOS since this OS is truly not ready for desktop class work.
What is "desktop class work"? I suspect you what you are generalizing to here is classic "high powered, multi window" app work. However, desktop itself is more a location and/or form factor adjective. The computer that is on a desktop does work for a user.
In the latter case, iOS can fit the bill. Just like increasingly growing number of folks can use what is actually a laptop sitting on a desktop to do "desktop" work now versus 10 years ago.
iOS restrictions of "one app window at a time" and one app at a time are more so driven by hardware restrictions: not too much computation load to drain battery and limited screen size. If ARM improvements get to point can run two apps at same time at same power load as old implementation could run just one and crank up the average screen size and .... ta-da there is a lot less need for iOS to hold onto the original restrictions. Apple has incrementally added more mutlitasking to iOS over time. In part, that was easy because it always was there and far more just an easing of restricted access to resources that were already being managed.
If this is a rumor to become true, I am sorry for Apple's decision: BAD MOVE.
Given the uptick in adoption of Chromebooks/Chromeboxes and the continued movement of average WinPC selling price down ...... it would be a very good move.
One of the growing segments is folks who need a computer that just does a limited set of stuff. Their workloads are limited. Hence price, not particularly performance, is the primary factor. If this workload is covered by iOS apps (e.g., email , browsing , video chat , streaming , etc. ) there isn't much the OS X variants would differentiate on.
When netbooks were 'hot' Apple countered with the iPad. Both were aimed at limited app workloads.
There were doubters about whether there was a real market for iPads because they couldn't do "desktop class" workloads. An iOS laptop form factor is likely in similar situation.
But again, I don't know how long they will keep giving up OS X for free and paying for the development from the hardware's no so deep pocket when they are reducing prices.
Who says they are reducing Mac prices? OS X isn't free. The price of OS X is just bundled with the hardware. Nobody is just buying hardware ( there are no offerings of OS-less hardware from Apple). People buy computer systems from Apple, not just hardware. They are buying hardware + software.
How long can bundling pay? If Macs get back on a growth path and keep their ~30% profit margins, then for a long time. The current fees rolling in pay to keep the lights on and developers paid. Where the "free" runs into a brick wall is when the growth curve goes significantly negative ( not flat or minor single digit negative blips). It can be run close to a Ponzi scheme or a bit more conservatively by taking some money out of the initial payment and setting it aside to offset "back porting" work to older hardware.
Apple cuts off older hardware over time so don't have to set aside huge amounts for long term "back porting" expenses. 5-6 years downstream the hardware will typically be cut off from upgrades.
The Mac market has grown large enough that the OS X expenses are amortized over tens of millions of Macs.
Even though they use open source to back up the main structure of the OS X, which reduces the cost of the development a lot, I did not see in BSD's page something that would indicate Apple wild move in this direction.
There is little to no reason for Apple to merge their updates into the BSD open source mainline. For a relased product there is more, but for some lab project ..... Apple isn't going to push the bug count down to release quality levels. Apple can fold bug fixes in the general BSD tree into their own private source code repository.
They would no be hiding it that much since they need to evolute a ***** load of drivers and others to hitch ride on current hardware available on the market.
Don't need a bucketload of drivers to get prototypes up and running on a small fixed set of hardware. And frankly, for iOS there won't be a bucketload of 3rd party hardware to hook too in a "desktop"/"laptop" context.