The task being preformed is not dimensional, in almost all cases. Computers are, largely, for data input and visualization. Data input is almost exclusively text-based, as the mouse is hellaciously slow. There have been a couple forays into true 3d interaction, but they've been horrible to use.

Visualization is already in pseudo three dimensions. Expensive displays can be purchased that allow true 3d.

This is an interesting concept. I think that if you were going to use a 3D system, then you should have a truely 3D input device. In other words, your "mouse" should actually move in three dimensions. Instead of just sliding around on the table, it would need to be able to move up and down too. The conventional mouse design wouldn't be appropriate for this. Maybe something more like a wand would be better.

The problem with using 3D is that objects in the back can get hidden by objects in the front. If you open multiple windows in MS Windows you always have this problem. A 3D system could allow you to theoretically use the windows in the back, but this isn't practical if you can't see them.

BTW, my mouse is optical and has a wheel but no balls at all. I guess I should call it a she.

Yep, you definately have a girl mouse...

The problem with the windows over-lapping is exactly what 3D avoids, as your viewpoint chages moving through the Z axis. I like the wand idea, only issue would be costs hence why I went for the mouse.

In response to toastydeath, yes, data input is the mainstay yet a 3D OS will not be to it's detriment, if anything quite the opposite. A lot of work done via PC is design based, 3D, architecture, website creation and otherwise. 3DOS would definately come into it's own here and show serious benefits.

I don't know about anyone else, but I personally feel the IT market has not progressed as much as it should do for a long time. I mean, really, we went through so many creations in the 20th century until 1980-90, around that point the wheels seemed to slow down and nothing much happened.

One reason I feel that we have not made such progression is Windows, it's truly archaic! The most recent one (Vista) is even based on a system over a decade old, it's just glorified XP.

...back to the point...

A true 3D OS would increase productivity and interaction. I would personally love for a web space to be a true 3D webspace, it's far from impossible...

About that mouse wheel... you can use it to control the Z axis. No new mice, just program the OS to make use of it.

How do you solve the overlap problem? You still have to use a screen. Moving through the Z axis only happens when you make it happen. Until you actually move, you can't see what is behind whatever's in the front. If you move to the back, then you can't see what's up front. If you make the objects up front small enough to see around, then they will be too small to work with. If you make the foreground objects semi-transparent, it will also make them hard to work with.

Moving "in and out" through windows with the scroll wheel is perfectly possible right now using existing OS. You just have to have mouse software that enables the scroll wheel to change the open window. Personally, I wouldn't want my scroll wheel to do that. I prefer being able to scroll up and down the page that I'm reading.

Apoloies, I may have mentioned mouse wheel as appose to mouse ball, which is unfortunately leading you down the wrong track slightly.

Furthermore, I realised a serious error with the sytem...

The interface device would have to have the equivelent of 3 mouse balls, one for perspective angle (which direction you are looking in), one for XY axis and one for YZ

Actually, I don't see any place where you said mouse wheel. You really need to explain how to get around the problem of hidden objects. I see that you have suggested changing perspective, but that still requires a separate action by the user that would be as troublesome as moving in and out to see what's in back. Though I have to admit, being able to change perspective would be cool! I would think that 2 mouse balls would suffice. One ball moves the cursor like normal, the second scrolls up and down by rolling it forward and back, and changes perspective side to side. This wouldn't allow changing perspective up and down though, unless you give up the normal scrolling function.

BTW, some design software does allow you to turn objects and change perspective using conventional OS.

Yes, I've just talked about the idea with my brother in law, whom is a proggrammer by trade. He stated that it would not take much to create a 3d GUI in director or something of the likes. Also having the day to ponder on the mouse, I thought that the best solution (other than expensive touch pads and wands) would be a single mouse with mini joystick under middle finger, (much like some laptops), to operate YZ position, a track ball under the thumb for XY position and the main optical sensor on the base would control view point direction/orientaition.

Semi transparent windows could be an option, but personally I think there's no real need to waste processor usage. Also, in a 3D OS you may get lost, so if the OS simulated a boxed enviroment, each corner or wall could be colour coded throwing dispersed light...

I'm not sure if this isn't just a solution seeking a problem. There have been a number of projects involving '3D Desktops', but for the most part, they seem to not offer anything new and useful. Just different ways of switching between application windows. What would benefit from a 3D operating environment? How would it look? How would a user interact with it?

I'm not convinced that we have a need of a 3D operating environment yet. Certainly we can make use of 3D accelerators for some fun eyecandy and compositing effects, but as long as we're stuck with our current desktop metaphor, 3D environments will likely be more trouble than they're worth.

I think we first need to rethink how we interact with computers. Maybe switch from application oriented environments to something more data oriented. But for now, leave 3D for the applications that really need them.