The physical form factor here described is "convertible slate". However, as someone who has spent a lot of their computer time using convertible slates (starting with a Compaq Concerto and a pre-release Dauphin DTR-1 back in 1992: yes, I was using "Windows 3.1 for Pen"), I will point out that with all of the circuitry in the slate the keyboard has to end up irritatingly poorly/awkwardly weighted (as it is way too light to make holding up the screen make sense) and is also bothersome to bring you (its "yet another fidgety component").
So, I then look at my 11" MacBook Air, and frankly: the dimensions are almost the same as the original iPad. It is /slightly/ wider (due to the constraints of having a full-size physical keyboard with keys that recess and also various ports such as USB that require a certain amount of clearance inside of the device, which is already so thin that these ports barely can exist in the device at all), but if USB is supplanted by something like Apple's Thunderbolt (convenient, huh), they could make it slightly narrower and slightly thinner.
I then argue what you're really going to end up with is a device more like the NEC Versa (1995-ish) or ThinkPad X61 (more devices I've owned), where the screen rotates around and folds back down backwards. This way, you always have the keyboard with you, and it can "take the load off" the screen part by having real circuitry in it (thereby giving it enough heft to balance the screen easily), which in turn allows the combined size to be smaller (as otherwise you are having to artificially increase the heft of the keyboard).
Apple, in fact, has been looking at such designs. We see patents from them for various ways of building a convertible slate for a while now. The original design had the screen hinged along the edge of the keyboard, where it could kind of slide down the case towards the end, and then fold back (away from you, as opposed to towards you) onto the keyboard, turning into a slate (which has the property that the screen's orientation doesn't change during the operation), but in 2011 we see the more traditional "rotate".
The 2011 design also makes it clear that this device could have a cellular modem in it, which means that were it to exist it would compete with a 3G iPad. That brings us to the listed constraint of "run apps written for iOS". Truly, however: much like you don't actually want a case that can house a keyboard and attach it to an iPad, you also don't want to be able to run iOS applications; the mistake in both is to take what you have now, assume "I can't have it both ways", and then come up with some kind of band-aid solution.
What you really want is to have a single set of applications that work well on either form factor due to a set of unified interface primitives. When you "convert" between a laptop and a slate, you don't want to be running a different web browser and a different word processor: you want there to be a seamless set of rather subtle changes to the apps you were just running (if any changes are required at all: in an ideal world you would want no changes, but there are practical concerns), allowing you to maintain mental state.
Again, this is the direction Apple is headed. In Lion they have started moving desktop Mac OS X to a world of full-screen applications, just like on iOS, and are borrowing many of the UI elements. They reversed the scrolling direction (to make the gestures common) and removed the scrollbars (which already aren't present on iOS for various reasons). Gestures are now permeating more of the applications, and with just another couple years of this, the difference is going to be quite slight; and yes: they have another patent on it.
Really, the only thing left would be to figure out how to unify the App Store experience. Ideally (for Apple), the resulting ecosystem would be quite similar to the situation on iOS: a closed store that Apple controls, with careful "security" measures at all levels to keep the user from "messing up" their device with software that wasn't centrally vetted. This would be accomplished with a combination of protected firmware and sandboxing of all applications, each of which would be isolated to their own state.
Well, with the advent of the Mac App Store we see Apple offering the same unified experience, and with Lion we saw them adding sandboxes. These sandboxes eventually became a required part of the workflow, and with Mountain Lion they are getting tighter. Meanwhile, Mountain Lion has added the "only from App Store" switch: defaulted to off (for now), but with another half-way setting, "only from registered developer" defaulted to on; even I felt this was ahead of schedule ;P. Again: we only need to look forward another couple years.
tl;dr I believe Apple's behaviors (and patents) agree: their goal is to provide this.
Another form factor option would be a Macbook Air with a double-sided display. Rather than flipping a screen around, you just close the lid to enter iPad mode, and open it for laptop mode.
Seamless app switching from iOS to OSX mode basically requires writing two different views, based on the same model, using a MVC design. That'll be easier for some apps (Safari, Calendar), difficult for some (Numbers, OmniGraffle), and maybe impossible for others (Xcode). It wouldn't make sense to try and force OSX and iOS versions of "everything" as some apps work best in one environment and not the other.
Windows 8 is half-doing this (from what I've read), and sounds clunky, but the verdict is still out on that.
Some applications are going to feel stuck in one interface or the other, but it is already the case that some applications already are mostly mouse and others are mostly keyboard: the touch screen is just another interface that is available, and for many applications it is in fact totally doable to make the transition between the touch screen and the keyboard as seamless as moving your hands from the keyboard to the mouse: it would be awkward to expect to see two views of the app, one mouse-driven and the other keyboard-driven, being flipped between via an MVC paradigm.
it would be awkward to expect to see two views of the app, one mouse-driven and the other keyboard-driven, being flipped between via an MVC paradigm.
I'd expect that both views would be fairly similar, like the two different views on an iPad vs. iPhone app, or when rotating 90 degrees. The OSX view would be tailored just for it; larger screen size, multiple displays, different input, etc... Apple is already unifying the look-and-feel of OSX and iOS apps, getting ready for some sort of merge (iCal/Calendar, Pages, Reminders in 10.8).
We're probably going to see five major input devices in all computers a few years from now. Keyboard, mouse/touchpad, touchscreen, voice, and physical gesture (e.g. Kinect, eye tracking). No piece of software has to use all of them, just as many as necessary. So I wouldn't expect "every" app to have an OSX and iOS mode.
In the future, a sixth input device will probably read our thoughts directly. :)
Where to run iOS, the "tablet" part would require a small processor, but the main CPU and circuitry could live with the keyboard. Someone was working on something like this with a stripped down Linux (or Android?) for the tablet, but I can't remember who (Lenovo?).
(FWIW, both Apple and Intel's website use the term "technical collaboration" to describe the relationship between Apple and Intel with regards to Thunderbold, which certainly seems stronger than "was the first vendor to license".)
> Again: we only need to look forward another couple years.
Two years is a very long time in tech. By then Windows 8 and a wide range of hardware it enables will be in full swing, not to mention Ubuntu on Android.
Also much of what you said is still speculation, since Apple seems to be the opinions that allowing OS X apps on the iDevices will made the experience and battery life worse. OS X applications expect to be running all the time and multitask unlike App Store apps.
They talked about how the post-PC world of iDevices will relegate PCs to be the trucks of computing. So their strategy seem set in stone for a while and they're selling boatloads of iPads, so I doubt they're going to pivot anytime soon.
Of course: anything said about the future is "speculation"; stating as much is pointless, as it is impossible to say otherwise. ;P We can only speak of the trajectory of things we can observe, such as market behaviors or patent filings.
To comment on "OS X apps on [] iDevices", that is still a notion that comes from the (I argue flawed) way of thinking that extrapolates a few isolated variables from today without taking into consideration how they will affect each other.
In this case, the performance and battery tradeoffs of multi-tasking are already changing in hardware: I do not feel like my 11" Air (again: nearly the same dimensions as the original iPad) is drastically worse for supporting multitasking.
Meanwhile, OS X /is/ moving in a direction of more restricted applications: it would not be surprising to see most normal software start being suspended when in the background, in exactly the same manner with the same exceptions as on iOS.
As for the "post-PC world", I fail to see how any of this is a "pivot": in fact, it seems to be exactly where they are going based on both their behavior (looking at the changes they are making to their product lines) and their patent filings.
Apple is explicitly bringing iOS features to OS X, where they are actively dropping Mac branding. Meanwhile, the performance of the iOS devices is increasing and they are bringing an ever more complete programming experience to it.
During all of this, they are actively evaluating hardware that combines the benefits of both of the platforms (specifically: keyboard and touch). Seriously: if I had a touchscreen on the MacBook Air, I'd question the entire point of having an iPad.
Given all of this, it would be shocking if they were not intending to unify the two platforms, and that is why this is relevant to this article: the author is stating that he doesn't want to have an iPad+keyboard, and it seems highly likely that Apple agrees.
(Finally, I fail to see why it is relevant to say that "two years is a long time in tech", and then bring up competition; it is certainly and obviously true that everyone is seeing this same vision, but the article was about people and their iPads.)
So, I then look at my 11" MacBook Air, and frankly: the dimensions are almost the same as the original iPad. It is /slightly/ wider (due to the constraints of having a full-size physical keyboard with keys that recess and also various ports such as USB that require a certain amount of clearance inside of the device, which is already so thin that these ports barely can exist in the device at all), but if USB is supplanted by something like Apple's Thunderbolt (convenient, huh), they could make it slightly narrower and slightly thinner.
I then argue what you're really going to end up with is a device more like the NEC Versa (1995-ish) or ThinkPad X61 (more devices I've owned), where the screen rotates around and folds back down backwards. This way, you always have the keyboard with you, and it can "take the load off" the screen part by having real circuitry in it (thereby giving it enough heft to balance the screen easily), which in turn allows the combined size to be smaller (as otherwise you are having to artificially increase the heft of the keyboard).
Apple, in fact, has been looking at such designs. We see patents from them for various ways of building a convertible slate for a while now. The original design had the screen hinged along the edge of the keyboard, where it could kind of slide down the case towards the end, and then fold back (away from you, as opposed to towards you) onto the keyboard, turning into a slate (which has the property that the screen's orientation doesn't change during the operation), but in 2011 we see the more traditional "rotate".
http://www.patentlyapple.com/patently-apple/2008/07/apple-re...
http://www.patentlyapple.com/patently-apple/2011/11/apple-wi...
The 2011 design also makes it clear that this device could have a cellular modem in it, which means that were it to exist it would compete with a 3G iPad. That brings us to the listed constraint of "run apps written for iOS". Truly, however: much like you don't actually want a case that can house a keyboard and attach it to an iPad, you also don't want to be able to run iOS applications; the mistake in both is to take what you have now, assume "I can't have it both ways", and then come up with some kind of band-aid solution.
What you really want is to have a single set of applications that work well on either form factor due to a set of unified interface primitives. When you "convert" between a laptop and a slate, you don't want to be running a different web browser and a different word processor: you want there to be a seamless set of rather subtle changes to the apps you were just running (if any changes are required at all: in an ideal world you would want no changes, but there are practical concerns), allowing you to maintain mental state.
Again, this is the direction Apple is headed. In Lion they have started moving desktop Mac OS X to a world of full-screen applications, just like on iOS, and are borrowing many of the UI elements. They reversed the scrolling direction (to make the gestures common) and removed the scrollbars (which already aren't present on iOS for various reasons). Gestures are now permeating more of the applications, and with just another couple years of this, the difference is going to be quite slight; and yes: they have another patent on it.
http://www.patentlyapple.com/patently-apple/2010/11/apple-wi...
Really, the only thing left would be to figure out how to unify the App Store experience. Ideally (for Apple), the resulting ecosystem would be quite similar to the situation on iOS: a closed store that Apple controls, with careful "security" measures at all levels to keep the user from "messing up" their device with software that wasn't centrally vetted. This would be accomplished with a combination of protected firmware and sandboxing of all applications, each of which would be isolated to their own state.
Well, with the advent of the Mac App Store we see Apple offering the same unified experience, and with Lion we saw them adding sandboxes. These sandboxes eventually became a required part of the workflow, and with Mountain Lion they are getting tighter. Meanwhile, Mountain Lion has added the "only from App Store" switch: defaulted to off (for now), but with another half-way setting, "only from registered developer" defaulted to on; even I felt this was ahead of schedule ;P. Again: we only need to look forward another couple years.
tl;dr I believe Apple's behaviors (and patents) agree: their goal is to provide this.