The desktop metaphor has been hanging around for a number of years now. The creativity within the software field should be able to produce a new cooler metaphor that doesn't involve dragging windows around all the time.
I believe that the desktop metaphor has soon been stretched to its maximum. What do you think?
Have you heard of any new cooler ways of using the power in today's computers that's different/more productive than the desktop? Or do you have a new crazy idea for this?
I think dragging windows around will be with us for a while, except it'll turn into "dragging objects around" a la Surface. All signs right now point to multitouch as the future.
Dragging windows will always be the way things work, unless we all go 100% CLI.
The next big thing will be multi touch monitors and table computers. But these will still conform to the "desktop" paradigm.
ps. subjective or argumentative
I remember seeing a talk where someone said that the great thing about language is that it allows us to communicate without pointing and grunting. If speech recognition ever becomes useful, I believe that it will cause a return of sorts to the command line. We may still point (finger, mouse, whatever) in order to indicate the noun that we are talking about, but we will get back some of the richness of language with indirect objects, adverbs, conjunctions, and so on, things that we already have in the command line but which are unavailable to most users who only know the GUI.
I can think of UI like the one shown in the 'Matrix' series, especially the third part viz. user are connected to the device and actually experience and feel their way thro' the program. The program reacts to the body and eye movement of the user. Very farflung though...
Heres my opinions:
Basically, I don't want to have to manage my computer at such a low level as we need to currently (nor do I want to wave my hands about Minority Report style, waaay too tiring). As a user of a computer, why do I care about filesystems, saving files, starting programs etc etc? I don't. I only care that my work isn't lost and that I can edit it and otherwise accomplish what I want. The OS should manage persisting my work for me. The OS should worry about what applications are running. I shouldn't even see an application start up, I should just see my document and as soon as I begin working on it.. I can. Without knowing that it had to run some program.
I like the interface in the movie "The Island" where they use a triangular prism like device in conjunction with a pen as the interface, where simply rotating or flipping the prism causes an action. Yes it is still "dragging windows around" but looks to be in a more intuitive and simpler fashion.
Interface metaphors will remain 2D until we have inputs and displays that aren't 2D.
As for "Minority Report"-style interfaces, I cringed really hard when I saw those "wave your hands in the air" interface scenes in that movie because I knew that people would latch onto that as some kind of wave of the future. That kind of interface is horrible for most things.
How many hours a day do you people spend on the computer? Eight, ten, twelve or more? Waving your hands around in the air is really, really tiring.
It's like everybody has forgotten the failure of light pens already. In the 70s and early 80s, everybody thought light pens were the ideal way to work with computers. They mimicked the pen-and-paper thing everybody already knew, and there was a direct 1:1 correlation between the movement of your hand and the "pointer." What wasn't to love? As it turns out, light pens were actually awful to use and the mouse became the dominant non-keyboard input device. They were physically tiring to use and required expensive displays.
Anyway, discussions over input devices and UI metaphors are really missing the point anyway. The problem is not that we can't point to things quickly enough, or that we need to do things in 3D, or any of that.
The problem is that we have too much information to sort through. How can a system - regardless of interface - help us to find the information most relevant to what we want? In other words, interface is important but the REAL issue isn't how we get stuff into the black box or out of it. The hard part is figuring out what to do inside of the black box, and that doesn't change whether you're doing things in two dimensions or fifty-seven dimensions.
I would say that the next real "metaphor" is multiple devices. Your virtual "desktop" space will consist of a variety of devices around you, in the same way that an extended desktop works in multiple monitors.
Your mobile device would be a viewport or an agent in the model represented by the UI, so that, for example, you will draw on your iPhone and get it to appear on your whiteboard or computer, etc.
Multitouch will become the major UI metaphor, but for power users and programmers, etc.. they will still use a keyboard. Maybe the mouse will be replaced, but probably no in the foreseeable future.
I think it's way to early to predict what will replace traditional GUI interfaces, especially because in some situations it's always going to be a good choice.
However, as we all know, traditional GUI interfaces aren't suited to many interactions that we invoke them on. Instead, other forms of interfaces will become prevalent in those areas, or even better, the interface will become lightly-seemed, transparent or even entirely natural. That is, the way we interact with computers is going to change dramatically and in a lot of different areas.
Multitouch seems to be the way of future, If we throw away the mice, we'll have multitouch screen closer to the keyboard and slanted at an angle and not straight-up. We could also have other (larger) screens on the walls.
Until we get a replacement for they keyboard which is much harder.
I'd like to say multitouch, but I really doubt it. Just because that's currently something new doesn't mean it'll catch on for anything other than specialised systems, eg kiosks and occasionally-used displays. If I had to use a multi-touch system all day long, you can guarantee I'd off to the doctor's with elbow problems after a year. And that assumes you can keep your arm from resting on the display and screwing with what you're trying to do with your fingers.
The best I think you'll get for a often-used touch display is to integrate a keyboard to it, and effectively turn the top half of your desk into a display where you interact more with the bottom section, occasionally using fingers to interact with the GUI at the top, perhaps organised like a draughtsman's table. I doubt it though, why spend that money when you can just use a mouse/keyboard/monitor.
I do think that the evolution of traditional displays will be javascript based, all the competing display technologies that have appeared recently may fade in favour of a single-codebase javascript that is hosted in a browser, or a desktop app, or your phone or TV.