Practibility of doing a 3D based OS?

General Discussions
#1

Ok, at first glance, that subject sounds/looks like a silly question. Why does a OS need to be 3D ? I mean, we have Windows, MacOSX, Linux etc., that supports everything we want, OpenGL, Direct3D etc. (And correct me if I’m wrong, MacOSX is basically all done via 3D?).

But what does it mean to do it in 3D? Well, for nearly 20 years now, we’ve seen computers and operating systems evolve from displaying only 1D things, like text, or simply lines, to displaying 2D, that is GUI applications. But what are the possibilities with 3D? It seems like a normal step to evolve the user-interface to the final level, being 3D. Current operating systems are 2D, trying to represent 3D. They’ve done a great job too. New GUI systems are now trying to implement more 3D interaction, like in Windows Vista, MacOSX etc.

What would a OS be like if it was mainly based on:

  • The low-level computer operating system control software
  • OpenGL graphics library
  • GUI system that thinks and does everything in 3D.

Is 2D enviroment dying out, soon? Is Windows just a relic of the 2D-era? Are we seeing a time where we laugh of 2D-GUI applications like we laugh now at text (command-line) enviroment?

Are we seeing a evolvement;
Up to 1995-1997: 2D GUI applications run in consoles.
2000: Console applications run in 2D GUI enviroment.
1995+: 3D applications run in 2D enviroment.
200?: 2D applications run in 3D enviroment?

Maybe there is a new unknown step yet to be taken in the computer evolution, just as nobody foresaw the success of GUI based operating systems. This might call for new thinking of how we interact with the computer, the 2D mouse (x,y) might become obsolete or/and evolve into something more, the keyboard will probably always be there.

Interesting to think about this.

#2

There is a reason to stick with 2D UI’s, and thats the simplicity of it, just think of a rotation in 2D and a rotation in 3d and you’ll see how complicated things get…

Also, another reason is that 3D stuff isn’t as powerful on a 2D screen, if it was a holographic sorta 3D image thing, then a 3D UI would be decent and 2D would simply not make sense there…

DP

#3

3d desktops are a waste of resources, a beautifull waste of resources.

in the end, you are using 2d uis on a 3d environment, like in project looking glass.

#4

The mouse moves only on a plane and todays screens are a 2D raster devices, therefore I guess 2D GUIs are most intuitiv (of course with exceptions, but in general). Innovative display and input technologies (stereo/volumetric, space mouse?/tracking systems) may change this, but that’s a long way IMHO.

Software is useless without the right hardware ::slight_smile:

#5

3D interfaces are a solution looking for a problem. The vast majority of computer users work almost exclusively with 2D data (text documents, images, etc) where adding third dimension offers no benefit.

Even if you have a use-case where 3d interfaces make sense (3D modelling, etc) the big problem is not 3D output -rendering technology is well understood- but 3D input. As anyone who has tried to do some 3D modelling and given up in frustration will know, manipulating 3D stuff with a 2D input system like a mouse is really, really annoying.
Until everyone has one of these, or possibly even a pair of these (also assuming the technology gets significantly less clunky-looking), 3D interfaces will always be difficult to use effectively.

#6

Theres a nice study for a 3D Desktop with physics simulation at youtube

#7

One immediate benefit I can see for 3d interfaces is that they give the potencial for a massive increase in screen real-estate.

#8

Yes, that’s true. Imagine being able to press 1 button, and you have 10x more screen area… although all the windows are small, it surely moves you away from the static resolution in 2D.

Well, of course this can be acheived in 2D, but better if you make use of 3D accel.

#9

OS X does this already with Expose, and Vista has an equivalent functionality. It’s great for sorting and managing a lot of windows, but it’s terrible for “increasing real estate”, the “increase” in real estate comes at a cost of resolution. Those tiny windows with the contents at 20% their full size are not easy to work with at all.

Besides, the 2D equivalent to increasing real estate (virtual desktops, or desktops that are of a larger size than the screen resolution and thus “scroll”, both very common in Unix environments) are no different from these 3D equivalents. They both push out more real estate from the monitor at some cost.

#10

The key is for each document to be visible at the resolution at which you desire it, at any given moment.

Look at the desk on which your computer is sat, is it covered in papers? mine is.
Can I see what is written on each piece of paper? obviously not.
But, I am aware of their presence, and if I so desire I can increase the resolution of it by bring it closer to my viewing portal - my eyes.

Having said that, digital paper is the real way forward, and it completely does away with the need for virtual simulation.
We already have a 3d universe, why bother simulating another?
We simply need to develop the mechanisms for integrating the digital processing capabilities of computers into the direct input methods that our bodies give us.

When I can spread 20 sheets of digital paper across my bed, each displaying the code of a different class - modify them via touch, transfer information between sheets, etc etc I will be a very happy man.